Great video. Saw your comment on Electronoobs' video on the DIY BTPS for under $10. Please consider an update to this video with some narration giving more detail. There are too many parts of this video that weren't obvious to someone just getting into this field. For example, some of the test you did with the multimeter and the modifications you made. There are many who'd be grateful for the effort and know what it takes to create content. Cheers
Hey, So this video is intended for a few different things. Firstly, all of my uploads have been an exploration of how the YT/Google search engine and suggested content systems work. All of the uploads have titles and descriptions in a number of different languages. The content is entirely void of spoken language for this reason. This stems from my previous ecommerce businesses and an in depth analysis of different sales platforms I've worked with. That was more of a personal curiosity. Secondly, the entirely intuitive based approach to this specific upload naturally limits the type of person that will find this useful and attempt it in practice. If someone needs a step by step and detailed walk through of this, I don't trust them to be capable of handling the dangers involved here. This kind of thing requires the person to have an intuitive sense of what is happening and the extreme dangers involved. This is very serious, and not something everyone should try. This stereotype probably does not encompass you, but it is reflective of my choice on the content I want to make. Thirdly, this is an oversimplified example of the similarities between a switching power supply and a linear regulator. People sometimes mention this when talking about SMPS theory, but I've never seen someone demonstrate it in practice like I did here. I understand far more about this subject now than I did when I made this. It's still a useful example to demonstrate the intended ideas, but it's not a super useful device in the first place. If you are just starting out, buy yourself a decent power supply for $50-$100 and go from there. This may seem like a big deal or too much to spend, but it is a savings in the long run. I'm literally working on my 6th (bad) bench supply design right now. I've learned a lot, but I can honestly say I would have been better off and would have saved more money if I had just bought a decent supply or two second hand on eBay. Alan over on his channel W2AEW has shown a couple of older HP supplies he likes to use and are available relatively cheaply second hand. I think I even saved the reference in my Fundamental Power Supply playlist. BTW, check my playlists if you haven't already. I don't use them as self promotion. I couldn't care less about self promo. Those are all of the references I save for myself and cover a massive amount of stuff I've come across and watched over the last few years. I've been trying to add info in the playlist descriptions about other related playlists. Everyone else can't see my alphabetical indexing system. At least, if you know it's there you might find what you're looking for. I need to subdivide the power supply playlist but haven't yet. It's got more than 100 references. I'm always adding stuff I come across. Playlists like this allow you to find much better references that the YT search system will not show you or feature. YT prioritizes monetized and recent content far more than relevant references. This stems from pop culture stupidity being the primary driving force for the platform. I made my references public because they can be useful for anyone following the same learning curve as me. Lastly, if you're looking for linear power supply designs, I've been looking at ones based on an old LM723 or CA723. Those are extremely precise regulators. They require a current pass transistor and a little more circuitry than a lm317 but I think they make a decent bench supply. You can find many circuit designs by searching images on duckduckgo or other engines for things like "LM723 power supply schematic." You can always just hit me up for specifics or a reference. I will share anything I know or where I go to find out more :-) Sincerely, -Jake
Wow, a lot to unpack in your reply. I read every word and totally get and appreciate where you're coming from. As for danger, we may be increasing it by not offering more information for those who will take risks regardless. On the other hand, we can't be held accountable for what others do with content of this nature. Ultimately, it's up to each content creator to decide what to offer. but you know all this. Cheers brother
Sir, I appreciate the comments you made about not wanting punter noobs who don't know which end of the soldering iron gets hot to mess with dangerous voltages, etc. I totally agree, but with one caveat: I totally agree with @cri8tor about the lack of information. I, myself, have been working with electronics since the 1960's, (when television flyback transformers generated 30KV+ and could knock you across a room!), and have a considerable body of experience, but I still found your video virtually incomprehensible. In my humble opinion an obscure and incomprehensible video is worse than none at all since people will try things, hoping to duplicate what you've done, and end up doing serious damage. I would really like to try to duplicate your results, but the lack of information is scary. Do you have other videos like this one - or better yet a web page that provides more detail - so that those of us who cannot read minds, ( :wink: ) can follow along? Thanks!
It works by modifying the voltage divider for the TL431 reference IC (voltage feedback) or something similar So basically the IC needs a 2.5V coming from the voltage divider, to trigger the voltage feedback pin So if you swap out the factory made voltage divider with a potentiometer, you can adjust the middle pin voltage, which goes to the TL431 reference IC The lowest voltage possible would be 3V, while the highest voltage depends on your SMPS circuit
There is one other factor. With the way these circuits are designed, the control chip must be powered by an auxiliary circuit on the primary side. There are a few ways this is done in practice, but most of the time there is a small set of extra windings connected on the primary side. This winding will usually have a single slow diode like a 1N4007 and a small capacitor of around 100uF@50V-ish. The quality of this circuit is largely determined by one simple thing, if there is a large-ish zener diode in this circuit. These chips usually have an input power voltage range of around 4.5 to 30 volts. This is the actual limiting factor when adjusting the output of a supply. It's this voltage range in combination with how the transformer was wound. Specifically, the turns ratio of all the windings is critical. When designing a switching converter circuit there are several factors that need to converge around the desired output voltage and current. These factors are permanently baked into the circuit. The windings' impedance, and the turns ratio of the Primary/Auxiliary/Secondary, and the switching frequency are all part of the design. They are not precision type elements though. Picture them like a wide bell curve. The output voltage, current, efficiency, and noise are optimized to somewhere near the top of the bell curve, but there is a lot of wiggle room on both sides of the curve. You can't make the peak or shape of the curve any different, but you can alter the parameters and have the output cross somewhere else on the bell. All of that said, when the voltage of the secondary side is changed, the voltage will also change on the auxiliary winding. A properly made power supply will also have a zener voltage clamp on the auxiliary power circuit. In most, if not all, countries a mains voltage switching supply must be able to have a failure on the secondary feedback circuit and not catch fire or go critical. Think about this for a second. If the feedback loop fails, that means the switching circuit on the primary side is going to push out the maximum dutycycle non-stop. So a supply must be able to survive this condition. The reverse aspect is the minimum voltage that the control chip can operate at. The chip doesn't require much current to work. This means that it's auxiliary winding can be quite small and insignificant. However, if the voltage is changed it will vary quite a bit. The single diode half wave rectifier doesn't help either. The way the Auxiliary winding is setup is the primary output voltage determination factor. In principal it should be wound to the middle of the chip's input voltage range. This will be a voltage close to the chip's lowest quiescent current. The Q current is a major factor in the no-load efficiency. Some people are cheap though and will split a thousandth of a penny making an aux winding that is as few turns as possible. If the aux winding has too low of a voltage the chip will have an under voltage shutdown function. This will make the output circuit cut out off and on. Most supplies can't go all the way down to 2.5v unless they were designed for close to it in the first place. If you know what you're doing and power the auxiliary circuit separately, then yeah, it will go down to 2.5 or below depending on the feedback resistor setup/mods. It will be noisy as hell though. It was designed for a pwm dutycycle optimized for the output, so the lower voltage makes the thing come on, go off, and take a nap while the output caps have an aneurism. Basically, it's not good to use this type of mod at the extremes, but if you have say a 14 volt supply at 100 watts and need a 12 volt supply at 80 watts this should work fine. It's not much different than a linear regulator. Don't hook up a 3.3 volt reg to a 12 volt input and expect to get the datasheet's max current figure. The magic smoke will find you. While there are more factors with smps design, it's still the same basic thing. -Jake
@@UpcycleElectronics Yes, I know, but it's only used as a voltage sensor, so most of the current doesn't flow through it. When the voltage divider middle pin reaches 2.5V, the circuit only sends a HIGH signal to the chip, which regulates the duty cycle accordingly. And the current through the 431 is very low, only a few mA. Also PC ATX power supplies that have KA7500 have a built in voltage reference, you just hook up a voltage divider (potentiometer) between the power supply output, and when you twist the potentiometer, the voltage on the middle pin changes, so the IC will always try to see 2.5V at its feedback pin, and so the output voltage changes. Also my power supply will rarely get current times when the load goes above 1A, and since my SMPS is rated 2A, it should be safe. I'm using this modified unit for my mini PCB drill, the motor current is ~ 250mA, occasionally it exceeds 1A, when I'm drilling something, but these currents are for short periods of time.
@@adamsucksatyt NO! A TL494/KA7500 is entirely different. It has 2 feedback loops. Both must be setup in tandem. Don't try to adjust a dual feedback chip. That's a whole different can of worms. Chips like the UC3842 have a much more simple current limiting circuit. The dual feedback chips must have a calculated design. I have several power bricks that were commercially sold as adjustable. A couple are just like what was in this video. Others are a UC3842 like circuit with an MC34063 on the secondary side with an adjustment divider in their feedback loop. I've never seen someone try to adjust a dual feedback loop controller. You don't want to mess with something like that either. Those multiple voltage rail supplies only regulate 1 rail's voltage and current. That primary rail must have the majority of the load on it or the supply will fail in short order. Leaving the other rails floating can cause problems too. Those are purpose built tools. They are not really flexible things to mess with... I was told this, and it still didn't stop me... so whatever :-) ...just saying... I built a box with 5 modified power bricks inside as a basic supply. If you know they are all isolated from mains, the outputs can be placed in series or parallel just like batteries. That's a much more flexible type of setup. Just be absolutely sure the supplies are totally isolated and the secondaries are floating.
@@UpcycleElectronics Dude calm down a bit, I don't know everything What do you think, does a 15 year old know all IC datasheets and make right assumptions? Also I'm talking about TL431, not TL494
@@adamsucksatyt ATX computer power supplies are almost always TL494 based. If you haven't found the channel Diodes Gone Wild go check out his stuff. He reverse engineers a ton of these types of supplies. I'm just learning this stuff too... although I have a few years head start. I've been saving references for awhile. They are public too. These are 2 playlists from the 6 I saved about power supplies. th-cam.com/play/PLPIwHuVy9EyPLzeyuEqKVK_LmWVQNzkZ3.html th-cam.com/play/PLPIwHuVy9EyPppZvFMXCheasl4DwlS2r6.html
Thankyou for the clear vid and how to do. I thought u did a great job without saying a word, was very clear to me anyway. So i collected lots of diff p supplies and had a go - i got 2 out of about 6 working. Also i found most didnt have just a simple divider to the TL. They were 2 resistors parallel and one series, combos like that. No matter which values or what combinations i removed and replaced with equiv value VR, i could never even adjust the output by more than a volt. However, with the couple i did get to work, like he stressed in the comments, the output voltage was unstable. Like, REALLY unstable the further down it goes from its designed voltage. I found once it got down near even half, it fluctuated up and down on my multimeter without even pausing, and that was without any load, ie, the output wasnt connected to anything at all, just multimeter measuring it. So i concluded, the ones i tried anyway, and with my very limited knowledge and skill, that doing this was not going to result in a new home made adjustable voltage power supply. Not something actually useful and reliable anyway, just limited use in its upper range. Near to what it was designed for. It was fun tho, and now im much more familiar with the little opto feedback to tl voltage ref section of these supplies. The supply that i had many of and thought would be the most useful when modded, i couldnt do at all. 24v, 10A, brand new boxed still, tl491, had as upper and lower resistive dividers something huge, like megohm or something, and the other so small my meter couldnt see it, like 1R or sub ohm, i cant really remember. I tried anyway, with closest VR's / pots i had, but no luck.
You had me impressed at first glance. Around 1:18 I felt your pain. If you use a very rounded chisel and a hammer you can easily remove the cover to the point that it can be easily glued back together. I'm very ocd and had to find a way lol
Rafael Roman A hammer and chisel could work. In a past life (about a decade ago) I owned a small autobody shop. I specialized in plastic and small panel repairs. Most plastic automotive panels are a similar type of ABS plastic as power supply enclosures. One of my specialties was plastic fusing and plastic reconstruction on rare panels and bumpers that are not reproduced by 3rd party manufacturers. This type of job uses a large variety of methods to reconstruct a surface. The most common and effective way to repair cracks or 2 ABS plastic surfaces you want to connect is with a soldering iron and some fine wire mesh.... ..Anyways the "just use an iron and melt it" idea is the same. These power supply enclosures are designed with a interlocking lip, 'tung and groove' style. The plastic is welded/melted together using an ultrasonic welding machine (welded using ultrasonic friction). This production process seems to be kept on the ragged edge minimum for the amount of effort needed to fuse the two halves of the enclosure. I have seen several that easily come apart with a few minutes of light persuasion. The trick is to be patient with channel locks or a soft jaw vice. Only apply enough pressure to flex the surface on one side of the enclosure while gradually distributing the pressure all the way around the enclosure. It will usually take 5 to 10 minutes of time to completely work the thing open. If you do it this way you will get a nice little project enclosure for free, about 9 out of 10 times. The edges will have some small imperfections, but they are not very noticeable if your patient while opening the thing. When I reuse one of these little enclosures I use one of 3 options. One option is to use a soldering iron to embed a nut in the plastic on one side, adding enough extra ABS plastic from something else I have around to thoroughly bury the nut. Then I drill a hole through the other side and bolt it together. Another option is to use hot glue to join the two halves as this is fairly easy to reverse if you really need to. The last option is to use a soldering iron to reweld the two halves together by remelting the flange. This is a permanent fix though. When you melt ABS plastic with a small heat source like a soldering iron, it tends to get more brittle and it loses some of its functional properties. Once this is done it gets harder and harder to rework. I learned the hard way a long time ago, you only get one chance to rework and get it right. Just a bit of my experience from ages ago. ....Alternatively, I recommend not being cheap like me and buying enclosures for a few bucks instead of wasting your time hording such useless things. I have 13 of them in my drawer right now ;/ The 'Holy Grail' of power supplies are the ones that bolt together from the manufacturer. They usually have interesting components inside as well. I even have a couple of wall warts that are held together with screws. The other 'extremely (functionaly) valuable' SMPS power bricks are the HP type with the 3 terminal plug on the DC side. Most HP supplies have the 3 terminal plug, but only about 1/3rd actually use it. The label will tell you, or if you look at the plug end just look to see if it has all 3 pins populated inside the plug end. This type of supply is dual polarity. They achieve this simply by using multiple secondary windings on the transformer. So the feedback circuit is just like the one I showed here. I tried to build one a few weeks ago to see if it is the same and it works fine. I can adjust that little supply from 12 volts to 20 volts with a positive AND completely isolated negative polarity. I can also wire them in series for 24volts to 40 volts. The circuit is earth mains referenced from HP so you must disconnect this wire, (actually a whole copper PCB plane is used) in order to isolate the power supply circuit. The one I modded is rated at +32volts at 1100mA and +16volts at 1600mA on the label. I think they must show a positive voltage on both channels on the label...cause people are stupid... but yeah, it's +/-16volts from the factory. I've tried to mod some other dual polarity SMPS designs like the old Dell's but they are super complicated PWM units that have multiple feedback loops. -All the best, -Jake ;)
Upcycle Electronics thanks for taking the time for that extensive reply. I have a box of 12v 5Achargers and I mean I have almost 100 so when I saw this video I had to see how it was done :) I wonder if you where able to pump 20v on only 1.6A supply, could I get a hover voltage with stable current with a bigger supply?
Rafael Roman I don't get out much after a broken neck 3.5 years ago. I tend to be a bit too chatty at times ;) Inductive loads are tricky. About half of the supplies I've tried get motors moving. Usually a large Capacitor can help get them started. A vacuum might be a challenge. Usually the generic supplies seem to handle motors and such a little better. I use one branded "LITEON" 12v 2.67A to run an old skool Milwaukee hand drill (1st gen cordless) and a small dust buster Black and Decker hand vac. It works as well as the batteries if not better... and it's unmodified. I've tried others rated at 2-3 times that many amps with no luck. I've even tried using a NTC to provide the start-up resistance, but even that failed. I think the load trips some part of the protection circuit on some chips, but I'm not positive. I should probably take apart the ones that work and some that don't and put them on a scope to figure it out, but I'm slow to do things on my work bench especially lately. Maybe I'll get around to that one of these days.
Hey man!! Firstly just wanted to say thank you for making this content, and especially then spending so much time in the comments helping/answering questions. I wanted to make the same complaint that a load of people already did haha (please explain more etc.) but I actually read far enough in the comments to realise why you made the video as you did and I appreciate the logic. (You could pin the comment which specifically mentions this, as a lot of people won't get that far down before they ask you the same thing!) So anyway I'm just at the beginning of this learning journey and yes this is quite a lot too much for me to understand at the moment, no problems I can deal with my frustrations for now, take a step back and I'm sure that in some weeks or months or even years haha I'll come back to this and find it more fascinating, accesible and useful!! I did open a laptop supply I have which malfunctioned a few years ago, fully aware that I don't understand it very well and that it's dangerous. I looked at the voltages on the multimeter across the various capacitors and watched them slowly going down which was cool (so I guess it has a bleed resistor for safety as it should right??) I then put a load across and watched them empty, and something I didn't expect or understand happened which is that after they drained, they slowly started regaining voltage a couple of mA a second. So today I will look up this phenomenon and try to learn about it. I also learned why the power supply failed, I think, which is that two of the resistors are a long way from their stated values - something else to go and look up (the list is looong)!! And what I have learned from here is that I think I will buy a cheapish lab power supply and make a DIY one just for a project, but a little further down the line, when I can understand it better. So again thanks for your help, I have subscribed to your channel and hope to be able to make sense of such videos in the future, and feel confident about safety and making useful and effecttive mods myself one day.
Thanks for reading a little deeper into the comments. I pinned one of the ones where I replied with my reasons for not talking in uploads. The capacitor reacquiring charge is a common thing. I've read about it, before but don't recall what it's called from memory right now. I'm currently on a tangent learning about digital electronics, linux, networking, and more specifically trying to wrap my head around how a UEFI bootloader system works. So, analog electronics are on the mental back burner. You also mentioned measuring resistors. Be very aware that it is virtually impossible to measure resistors and capacitors when they are in circuit. The device making the measurements is using ohms law. This means it is applying small voltages and/or currents to detect changes between the probes. As a general rule of thumb, unpowered circuit boards are ravenous beasts willing to devour every spare electron they can acquire. When testing a component in circuit, you need to have known conditions, if you want anything accurate or even relevant. This is why, in the upload, IIRC, I either removed and measured the resistors or I used the numbers printed on them to state the values. I used continuity in circuit just to figure out what was connected to where. I had to measure the voltage divider while under power in order to figure out the power, ground, and voltage reference node. I don't think I ever showed a resistance measurement in circuit. If I did, I apologize, as that would be pointless. There are ways of measuring some of these components in circuit. The tool used is called a LCR. It measures using frequency instead of using simple voltage/current relationships. The LCR is actually measuring impedance (resistance@frequency of X // aka AC Resistance). Don't worry about that too much though. Regular ohms law type stuff is much more important to grasp and use. The LCR stuff is really only useful/worth the investment for people that understand and can manually design radio circuits and frequency filters from scratch. Most of us will never get very deep into that can of worms. You could spend a lifetime going down that rabbit hole and still not fully understand the subject. In other words, it's not a path you should wander down randomly. That should be a well planned trip with lots and lots of planning in advance. Most people will find and use premade circuit elements when they need a filter circuit. You'll wind up buying components with specific values, and from reputable sources so that you don't need the expensive tools to make the LCR measurements yourself. Lastly, resistors don't usually change values. When they fail it's usually something else that caused a lot of current to move across them and they burn. You'll KNOW when they fail. It's that big charred black mess. Usually, switching supplies fail due to capacitors or the switching transistor. This is complicated. If there is not a snubber network on the mains side, there will be a voltage spike on the main switching transistor that can kill it with normal use. Transient voltage spikes on the mains power line in your area are also common (everywhere). MOV's protect against this, but most cheap PSU's do not have them. The mains rectifier diodes are also a common failure mode. Bad caps are the most common failure mode though. When they fail it usually kills the controller chip. Get The Art of Electronics 3rd Edition. Search for the PDF online. Uni students pirate scan and post them to combat ridiculous book prices. See if it's a good fit for you before buying it (yes you need the 3rd edition). Also get a copy of Hacking the Xbox by "Bunnie." Second-hand copies are dirt cheap on eBay. First, it isn't about the Xbox. Second, your interest level in the xbox/gaming is irrelevant to the value of the book. Those are my 2 most useful "general references." I still don't have a copy of the ARRL, but that's another good one. I've gotten library copies multiple times. Hopefully this helps :-) -Jake
@@UpcycleElectronics Hey Jake thank you so much for taking the time to reply and the suggestions and information. Funnily enough I've used linux for years and I remember trying to get my head around UEFI boot systems a few years ago - I don't think I got very far with the understanding part haha but enough to do what I needed. Well as with everything, the more you know the more you realize you don´t know, and evidently I'm right at the beginning of this journey. Feeling very motivated to learn more though! I managed to do a super easy 3d design on Tinkercad (kid's online CAD programme, it's awesome if you haven't heard of it), then I made all the pieces out of "plyboard" - using PVA and cardboard in a 3 way ply is suprisingly strong - and built the box. I rescued a broken 3w bluetooth speaker which sounded awful before, and managed to combine it with the solar panel and [6x] bigger battery from a solar charger which all just happened to be the same voltage/chemistry....and it sounds amazing and works really well. So you can imagine I got really enthusiastic, and have now taken apart a few broken bits and bobs and have a few projects in mind. Inevitably I am now realising that to do anything mildly custom, which isn't just connecting up pre-made modules, there is actually a huge basis of knowledge which I need to chip away at if I want to slowly build a decent understanding. Yesterday I tried to measure the amperage, in the same way as one does the voltage, and couldn't understand why I shorted the circuit lol. Then of course I thought about it a lot, and went and watched a simple tutorial, and everything became so obvious. Fortunately all low amp/voltage things so no damage to the multimeter, and after the obvious homer simpson "doh" moment and feeling stupid, it actually felt great to make sense of the obvious. Today you've given me another one of those moments. I was kind of thinking about it yesterday as I was measuring the resistors, but obviously didn't make the click. Today I bought a DIY kit for a bench power supply, and I think it might end up sat on my shelf for a fair while while I get constructing some super simple circuits and do some more studying, but it's a nice goal to aim towards and would be fundamental for a couple of things I'd like to do. I have a copy of "Make: Electronics" which I am going to get started on, and I will check out the other books you recommended too. Good luck with your various endeavours and if I get a channel up at some point to have a look at any of these projects (I think the speaker idea for example, would actually be a cool learning project for other people too) I'll let you know!! Alternatively if I never comment again, we can assume that I didn't fully understand something about high volts at lots of amps and touching bare wires or something like that lol Thanks again man!!
@@TheKoodus Hey, if you ever get around to making some content, I am your first subscriber :-) Electronics is a deep hobby. If you go down this rabbit hole you may find it's a lifer. I started into the hobby after a car hit me bad, like really bad, on a bicycle commute to work Feb 26th 2014. No sympathy needed. Just saying, I started into hobby electronics thinking it'd help me burn a few months to a year while I recovered from a broken neck. I never fully recovered physically, so I've been at this every day since and still feel like I know close to nothing. Best of luck :-) -Jake
Brilliant, I have a couple of unused laptop adapters laying around and was thinking of disposing them, but you've just given me an idea on how to re-purpose it. You just earned a sub :D
No. An ATX power supply uses a different type of feedback loop to monitor the voltage and current actively on the main secondary rail. The circuits shown here use a simple resistor to sense the current on the primary side while directly sensing the output voltage of the secondary. This single feedback loop is much easier to modify. If you modify an ATX supply like this it will fail very quickly. That type of design must be engineered for a specific application. There are channels that go into designing circuits like an ATX power supply. Mr Johnnny and Diodes Gone Wild are both good channels for this type of content if your looking to build something powerful. The ATX circuits are usually based on a TL494 IC. That IC is also called a KA7500 that is basically the same thing and is pin compatible. If you haven't designed SMPS circuits before, these designs are very complicated to do. I would start with something like a UC3842 IC design if you really want to build a mains powered SMPS. If you've never built a SMPS before start with the MC34063 buck converter style design. That's the IC used in most older phone chargers for cars (your car is 12v and your phone is 5v so these are simple buck converters). The MC34063 is only capable of supplying around 500mA but it will start you on the path to SMPS design. Danyk is the content creator for the channel Diodes Gone Wild, if you search the internet for "circuit" or "schematic" and "Danyk" you will find his tutorials about UC3842 SMPS designs. That's probably the easiest place to start. -Jake
thank you very much , I,m new in electronics but I read more, and yes I subscribed all these channels you mentioned I want to re-engineer control chip in Atx modifying TL494 and protection lm393 with 30v and 5A, can I use Reverse engineering by clone PCB and convert it to Gerber then get the schematic (the secret) ? and by the way, I love what he did: www.instructables.com/id/Adjustable-High-Current-Lab-SMPS-Out-of-a-Standard/ best wishes : )
The key is to understand that Ohms law always applies. Watts = Volts * Amps. The current/watts rating is built into every part of the circuit. The size of the magnetics, the gap between the two halves, the size of the copper wire used, and the number of turns, are the biggest factors. The diodes, and switching transistors are also selected for a specific output current. When you change the output voltage, you're detuning the circuit. In other words, you should derate the output. If you change the output voltage by 20%, you shouldn't expect more than 80% at the output. For example, let's say the supply is rated for 10 volts and 50 watts. If you change it to 12 volts, at best you still have a 50 watt supply. You've changed the specs by 20%, so the sensible thing to do is assume a 20% reduction in the maximum output. This means your 10 volt 50 watt supply was capable of 5 Amps max. At 12 volts 50 watts is 4.1 Amps. If you are sensible and derate the power supply by the rate of change, you wouldn't use it for more than 3.3 amps. I didn't show "how to hack a switching power supply." I showed how to make a little adjustable bench supply. A bench supply is something that is always attended and generally gets light use. If you have a purpose in mind, than this is not a good project for you. When designing a switching power supply there are around a dozen factors that all come together at one point, the desired voltage/current output at a specific switching frequency. This is like the peak at the top of a hill. There's lots of slope on both sides of the hill, but you aren't going to make a bigger hill.
@@UpcycleElectronics Hello dear youtube expert ! thanks for your detailed answer, i got what you meant by maximum, but i thought that through some fans and increased capacitor, i could increase the output current by 30%. On another hand, why did you put TWO potentiometers ? i could thanks to you better (not fully yet) understand the voltage divider thing, but why making two variable resistors since the max stays the same : 20Vdc.
@@hichamtassi1753 If you watch the video carefully you will see what is happening with the 2 potentiometers. One is a trimmer only. Once it is set you never touch it again. The second pot is the adjustment setting you use all the time. If you watch carefully I showed how the pots were adjusted. I then connected the power and set the maximum voltage level using the trimmer pot.
@@UpcycleElectronics Ok, perfect then, i tried to change the current through the PIN3 of the SMPS chip, i could only increase the output by 0.5A and it was not linear, so not worth it...
Multiple voltage tap types of supplies are usually much more complicated than what I showed here. These simple smps supplies only have 1 feedback loop. This makes them much easier to modify. If you're looking for a good source of circuits like a current source get the 3rd revision of 'The Art of Electronics.' If you look around there is usually a PDF floating around the web for "free." I save all the electronics references I come across in publicly listed playlists. This is my list for current sources and stuff: th-cam.com/play/PLPIwHuVy9EyOm6MjX1_J_YW9QKfQcI2Vg.html And this is my playlist for power supplies in general: th-cam.com/play/PLPIwHuVy9EyOtJ_LYw_Rc8oFvBd5TUTYN.html
hi..nice job I have a dell laptop adapter 19.5 volt and 3.35A. I want to get 12 volt and 3+A from this adapter can you please guide me I'm not expert in electronics please kindly guide simply if you can..thanks in advance
This example is for someone that is just making a little lab bench power supply. The video illustrates how I modify these kinds of power adapters. If you need a specific output for a project or device you need to purchase the appropriate product designed for the application. Never modify a charger's output like this and leave it unattended unless you know all of the math involved to calculate all of the potential failures in the power supply. If you don't understand what I have shown here, messing with power supplies like this is a very bad idea. Switch mode power supplies are designed for a specific output, there is a certain window of adjustment that will still result in an output voltage but it is compromising the design when a different output voltage is set. The modified output is barely useful for the most basic types of bench testing and experiments, but it is in no way a suitable replacement for a properly designed power supply at a specific power output.
The way you Opening adapter cover is wrong could be damage the rite way is just apply karocine oil aur sprite/ thinner arround the joint then tap gently with screw driver base all 4 side it will open the cover without breaking.
it is confused, make a note to every step that what you doing, it will be help for us .. anw this is first tutorial that I've seen increasing output voltage - thumbs up guys
TheJavaSync I tried to explain it in more detail in the video comments description. I am trying to create content that is as visually intuitive as possible. The title and basic comments are translated into several languages using Google Translate. I'm trying my best to make something that is useful to people who may or may not speak English.
TheJavaSync I just added to and modified the comments description to add the following: "I replace 2 resistors from the DC output side of the power supply. These 2 resistors create a simple voltage divider. The voltage divider is connected to a zener shunt regulator, and it's output goes to an optocoupler. You don't have to know or understand the circuit. The optocoupler is always a 4 pin dip chip that crosses the isolation barrier under the transformer. Usually there are two resistors for the voltage divider and some others for biasing the opto and/or the shunt regulator input voltage pin, don't worry about these. The shunt regulator itself is always a TL431 The shunt regulator will have one connection that has the full voltage output (or close to it(biased)), a second connection will be at ground, and the last connection will be at 2.4-2.5 volts(Vref). Just look at the output caps to find positive and ground traces. The simple equation for the feedback output is Vout= (1+R1/R2)*Vref I've shown an example with the values from this video. The equation came out to 12.44 volts. If you watch carefully I actually got 13.44 volts. This was because I didn't want to complicate things by showing how there was a small additional bias on the voltage input pin of the TL431. If you notice it wasn't at the full 12.12 volts of output power because of the extra resistors below the ones that I replaced with potentiometers. The only reason I mention the equation is to insure you don't blow up the output capacitor(s) accidentally. The accuracy of the equation doesn't matter, it's just an approximation. The resistors are easy to identify, R1 will be connected between the output voltage and Vref pins, R2 will be connected between the Vref pin and ground. Pick 2 potentiometers that are close to the values of these resistors. Resistor R2 should be a trimmer pot because you only need it for setting up the maximum output voltage. After you've set the maximum output, you use the potentiometer you replaced R1 with to adjust the output voltage. There is a limit to how high and low of an output voltage you can achieve with each SMPS design. This has to do with the transformer specs, frequency controller chip, and timing capacitor. Most power supplies that I've modified are stable from about 7.5 to 22ish volts. The supply can be biased higher or lower but when a load is applied it will get pulled down."
TheJavaSync I'm 4 weeks into the current project but I'm a slow partially disabled gimp. I may do another short side video on enclosure building/plastic welding though. The Next Project: I am waiting on a transistor tester someone ordered for me. I should get it in the next couple of days. I really want to demonstrate the difference between switch mode and linear regulators. I am trying to show how you can build an adjustable AC to DC power supply using an upcycled phone type car charger and a salvaged transformer. I want to be able to add a current pass transistor to the setup in the future, but I must test several phone type car charger circuits and their components first. I want to do all the math first to be sure the common components, most people can find, will be able to handle the design. I'm trying to avoid telling everyone how to do a massive amount of math to figure it out for themselves or give them a "paint by number" list of junk to buy. I want to distil the information down to its bare necessities first. I'm also using an analog Voltmeter upcycled from a cheap multimeter, and an ammeter upcycled from a car battery charger. I'll give references to how to reuse them. Once this supply is complete I'll be able to demonstrate how to reuse them myself, as this project is able to adjust down to zero volts or even below zero if you want. Judging by my current situation I should have this project up before next weekend so long as my back stays in decent shape ;)
hello, so if i would calculate R1 and R2 to get ~12V and replace those 2 SMD resistors with 2 other SMD resistors i would set it permanently to 12v without any limitations (if the power supply was i.e.19v) ? or do i need to be carefull if i draw more current?
12 Volt supplies are the most common ones I've found in general. I would use one of them personally... With a peak mode controller you'll be lowering the 2nd half of the wave basically making more current ripple noise. It might stress the capacitors more depending on how or if there is an LC output filter on them. To answer your question, yes, it's just a matter of calculating resistors. However the equation is somewhat simplified. I would recommend you to replace one resistor with a similar value potentiometer. Then use the potentiometer to dial in your desired output. If you want to use 2 resistors only, measure the potentiometer's resistance carefully and match it with a resistor. It just has to be a simple peak mode controller Circuit for this to work. If your trying to mod one like the old Dell laptop power bricks with well over 5 amps of current, you'll probably run into a PWM controller that is way more complicated. Those are about like ATX supplies. I have one of those. It has 2 transformers inside and is way too complicated for my tastes. It has 2 output polarities but it uses earth ground from the AC mains to create them and the circuit is really odd. Any way around it this is a mod you do for a "bench" power supply. This is not the kind of modification that you do to a device you plan to leave unattended. When a SMPS is engineered everything is designed to create an intersection between the oscillator timing and core saturation regions. If you go too far away from this crossing point in one direction(lower voltages), you'll get a lot of noise, if you go too far the other direction you'll get heat/smoke. All that said, here's a commercial example of how versatile the output of a switch mode supply can be: i1160.photobucket.com/albums/q500/jakebikereligion/20170602_182031_zps24vdy4hq.jpg
Love your video. Very clear steps. You know what you are doing ! :-) I'm trying to modify a Dell PA-6 Family. Do you have any experience with that model? Originally the Cathode and Ref of the AS431 are connected (Zero Ohm). I disconnected them and inserted voltage divider which increased the output voltage at the Cathode to 3V but the output of the power adapter is unstable. At power on, it goes to 18V and then slowly drops to almost 0V Any tips or ideas? PS: a schematic of the PA6 family adapters seems nowhere to be found thanks ! chrisV
Because this is essentially free. Both options are super noisy and pretty much useless for a real bench power supply. The little '0-30 volt 0-2amp' bench power supply kits based on the TL081 are a far better option to build. There is also a kit based on an ATmega8 with a little 1602 display. I have both and have been thinking about doing a video showing them in trace by trace detail along with modifications. Both kits are based on open source projects. The TL081 kit is from an old magazine article from the 1980's and the mega8 is published online. I am currently working on a huge AVR Transistor Tester project and don't want to get distracted though, so I don't know if I will ever get around to content about the power supply kits. The cheaper TL081's are probably the most interesting kit. If I understand the circuit correctly, the board uses a small zener voltage reference. This small voltage is then amplified to make a full scale voltage reference that goes all the way to the max output voltage. This full scale reference is then attenuated to create the desired output. That's completely different than most linear power supply circuits. I tried to model this circuit using the tiny "EveryCircuit" spice simulator. That app is too small to fit the whole circuit and it can't handle multiple op amp supply voltages as the circuit is designed. I've been meaning to try it in LT Spice but haven't yet. ...anyways... all I was showing with this upload is how the TL431 voltage reference is used to regulate the output just like a linear regulator's internal voltage reference works. There are problems with this circuit and it should never be left unattended. The biggest issue is how the auxiliary power supply for the switching IC changes when the secondary output is adjusted. The second biggest issue is that the frequency of the switching controller and the inductance of the transformer are specifically selected and designed for. Changing the output voltage also alters the crossover point of the design and will cause problems. The third biggest issue is that these supplies usually have an output filter with 2 capacitors and an inductor in addition to a choke. Altering the output characteristics will cause issues with this filter rendering it useless at best or even causing it to amplify the problems further. There is a margin of error built into these designs but the further the output gets from the original design the worse the problems will be. This applies to all switching power supplies including the Chinese modules. When it comes to a bench power supply the most important traits are stability, over/under voltages, and what happens when it is powered on or off. Most switching supplies suck at these specs. -Jake
Yeah, but sometimes opening a power circuit and plugging it in is not a good idea for begginers (and the input caps are also dangerous). Those who dont understand properly could got hurt or killed. And not all laptop charger has the same circuit, if someone actualy tried to modify their own charger, they might blew something up. Anyway, thank you for your explanation, keep up the good work sir! :) -sorry for my terrible english
I don't recall, and am not rewatching, but the first adjustment pot sets the high voltage limit. It's the one I gestured to show: "once it is set, don't touch it again." The second pot adjusts the output. The lowest voltage setting is something you can't control unless you (re)design the circuit. It has to do with the auxiliary coil on the mains voltage side. The aux winding ratio/rectification/cap are usually the main issue. When you adjust the output of the secondary, you're also adjusting the output of the auxiliary winding. Once this is low enough, the switching circuit loses power and the control chip shuts down. You need to completely understand the feedback circuit if you wish to modify the aux circuit as these two circuit blocks are integrated on many supplies. Watch a bunch of videos from DiodesGoneWild where Danyk explains switching power supply topologies in depth. That's the main place where I learned. There is a lot of potential for issues, and "beware of deadly mains voltages" should be quite obvious here, but that said, it is possible to use a second isolated power supply in place of the aux winding on most power bricks with a controller chip. In theory, this would allow you to adjust the supply down to whatever the control chip's internal reference voltage is. Also, I shouldn't need to tell people this, but many seem to need it: modifying the circuit as shown is intended for simple bench use. Never leave it unattended, or attempt to use this with high loads for extended periods of time. The original as-labeled output should also be considered "ideal." Derate the output by the percentage change. If you adjust to 1/2 the original output voltage, assume you've cut the efficiency by 1/2, doubled the noise, and halved the maximum output wattage. No matter what you do, the wattage rating can only go down. This limit is baked into every component used in the circuit design and can't be changed. For instance, a 10 watt supply is never going to be more than a 10 watt supply. If it is set at 10V, 10 watts is 1 amp. If the output is 20V, 10 watts is 500mA. You'll never be able to violate Ohm's Law. -Jake
@@UpcycleElectronics do i need to change the orginal value resistor ? Gnd to ref pin is 10k and ref pin to + is 60k+500ohms. All i need is to change the 60k resistor with 60k or 50k pot right?
@@smoke_stackz3168 If you do that, you won't be able to control the voltage as shown. The idea is to set your (main/second) adjustment pot to maximum resistance, then set the resistance of trimmer pot (the one to ground), as shown, until the maximum voltage (with an extra margin) for your capacitors is reached. If you skip this step, you're probably going to blow up your capacitors, and you should check the voltage ratings for everything on the secondary side too. You're probably going to burn/blow up everything as you haven't set the output maximum, but you might get lucky.
@@UpcycleElectronics ur right i replace the resistor 10k to 4.7k and replace the 60k with 20kpot+4.7k and the output is pulsing at 13.5 to 13.8v looks like the adapter is turning on and off and led is blinking slowly
@@smoke_stackz3168 Some times they need a load on the output to work at a little bit lower voltage. Usually this means the aux winding isn't making enough power though.
Thanks for the comment. Sorry but...no you can't make it more powerful. You need to learn how inductors work. You can't make the transformer's saturation region change by changing a resistor. The power will go down the further you adjust the output away from the original design. The power factor will also change, as will the efficiency. This mod is intended as a simple way to create an adjustable bench power supply. I don't recommend trying this for any power supply that you leave unattended. There is an open loop protection circuit on the primary side of all UL listed "off line" (AC mains) switching power supplies that should prevent the output from getting high enough to cause it to catch fire because of oversaturating the transformer. If you want to create a higher output it requires a lot of math and an understanding of high frequency transformer design. The supply you have will be designed with a specific turns ratio on the primary and secondary windings. This will also match the cross sectional area of the core. All of these calculations relate to the switching frequency set by a special capacitor and several other factors depending on the design of the switching IC. It gets really complicated to understand and work out all of the math. However after everything is calculated for a circuit like this, the designer of the circuit is not creating a precision device. The final circuit design has a considerable amount of wiggle room. I've simply shown how to take advantage of this "wiggle room" to create something useful. -Jake
FREAD Go to the app store and get an Ohm's law calculator app until you get a solid grasp of how Ohm's law works. Power(watts) = Voltage × Current (amperes) It's usually just written P=VI Watts, Current (Amps), Voltage, and Resistance are all related. If you have a circuit with a certain Voltage, and Current already, then you decide you want more voltage, no matter what you do to the voltage it is not going to add more energy into the circuit than what you started with. So if you can't add more energy, how do you change the voltage? Simple, you must reduce the current. This will always happen whether you expect it or not. If you have a power supply that is rated for 120 watts, that's the total energy available as that power supply is designed. You would need to totally redesign the entire circuit from scratch to make this number higher. That said, if your looking for a 240watt 19 volt supply, your looking for around 16 amps of current. That's getting into industrial grade hardware. A regular household circuit breaker can't even supply that much current. You'll need a design specific for that kind of output. Once a DC converter type power supply gets past around 100 watts the circuitry needs to incorporate power factor correction. PFC is complicated and I don't completely understand it myself. It's a circuit that prevents the supply current from lagging behind the peak voltage of the AC waveform. It has to do with efficiency and how much noise the circuit adds onto the mains AC power line. What I showed here will only work on simple peak mode type circuits anyways. If you open up that Dell you will probably find a completely different type of circuit. The type of circuit I showed here are the most common ones for laptops and printers and they usually are rated for 12 to 36 volts and 2 to 6 amps max. Once the current goes higher the design usually changes to a PWM type of controller. They are more flexible for high current. The main difference is that they have 2 voltage feedback circuits. There's one loop for the voltage and another for the current. You can't modify those with potentiometers because both feedback loops are connected in how they cross each other. If you just change one feedback loop, for example, the voltage feedback only, the circuit will fail quickly. This type of design is used in almost all PC power supplies like the ATX. When people try to make bench supplies with them they tend to fail because of how this feedback system works. They are much harder to modify and understand properly. The most common IC used for that type of supply is the TL494. There are many circuit references online for how to build a PSU around that IC. There are probably a few calculators available too if you search for "TL494 design calculator." That's the IC you will probably need to use or figure out if you want to build something as serious as your 240W@19V/16A setup. -Jake
Yeah. I've actually played around with this circuit again recently on my bench. I've been trying to figure out different combinations to create a better adjustment range. Just using a power NPN creates too much heat for the enclosure and heat sink. As an aside, there is usually a spot for a second TO-220 rectifier in most of the generic laptop PSU's. I have been playing with ways to make use of this by cutting a few traces. This allows use of the heat sink already installed, and the potential to keep everything in the original enclosure. The difficulty for me has been finding a way to track 2 separate adjustments, one for the SMPS and the other for the npn pass transistor. I've been playing with ideas. I think I need to use an op amp to effectively control both circuits with the appropriate offset. I think I need to order a small dual potentiometer though. I want to try a circuit using 2 separate resistor dividers and a zener to set a minimum threshold for the SMPS. However I have also done a good bit of experimentation with changing secondary windings and ratios on flybacks recently. The best SMPS hack may be adding a small additional secondary winding. Just a few turns is usually all anyone needs for a few volts. This winding can then be shunt regulated for stability and used to power an independent adjustment circuit using a couple of scavenged 431's, a power NPN and a N-FET. This can create an adjustable voltage and current. Unfortunately this probably won't fit in a factory laptop power brick unless you get super creative ;) -Jake
had a 4amp 16v supply i hacked it to 19v supply but iam getting lower current output and voltage fluctuating when load is connected what should i do to rectify this problem seems like its working frequency is reduced please help
Your drawing too much current then. Your probably saturating the transformer or tripping some kind of protection circuitry related to duty cycle. The ideal calculation of ohms law for the information you gave is simple, 16V@4A= 64W. If you reverse this you can say your power supply is capable of 64w when operated as it is designed. Next you can calculate the current of 19 volts at 64 watts just to get an idea of what the perfect world number is. That comes out to 3.3684 amps from 19 volts@64watts. Of course this will never actually happen because a SMPS is designed for a specific crossover point between the oscillator timing and the flux density characteristics of the transformer. The farther we deviate from the crossover point the worse things will get. Personally, I would assume I could achieve no more than 70%-80% of the ideal number. I specifically described how to build a bench power supply for simple bench testing in this demonstration. Never leave something like this unattended, especially if your using high current at a higher voltage. If the device is compliant with regulations in most countries the design should be limited by the optocoupler and circuitry on the primary side before it gets into a catastrophically dangerous condition. If the optocoupler or any part of the secondary feedback loop gets damaged, the output of the controller will peg out at maximum. The voltage reference inside the actual SMPS controller chip will be compared to a zero instead of the feedback voltage it normally "sees" from the primary side of the optocoupler. Any safe design should account for this condition. The largest physical limitation is the duty cycle of the controller and saturation of the transformer. If you want to push SMPS to its limits you really need to fully understand how it works. It's not simple. The transformer equations are probably the biggest challenge. That's actually where I'm at with my own learning. I'm not an Electrical Engineer, I'm a hobbyist like you. I read a lot and try to share what I've learned. I started a series on building one of the transistor tester clone tools because I need to measure inductors before I can take my understanding of SMPS circuits to the next level. I have been saving a lot of good references on the subject in my 'Power Supply' playlist on my channel. The hardcore SMPS stuff is towards the bottom. There's a few practical examples with transformer design/winding from Silvio de Leonardo, and the theory stuff from Sam Ben-Yaakov is mind blowing. He has a video posted on everything SMPS engineering related. You also need a thorough understanding of MOSFETs and especially their gate capacitance effects. I think the best possible reference is from the Free MIT Open Courseware channel, specifically the 6.002 classes and lecture 9 a and b of that course. I hope you understand I am not trying to discourage you, or talking down. I'm no expert. I just rea a lot and watch a bunch of technical stuff. These references are not something I'm telling you to do. They are literally what I am currently working on understanding myself. The latest SMPS project I got 3/4 of the way through before getting a component tester has seen me smoke 3 IC's, a trimmer pot, and a FET all in the name of exploration and learning. Be careful, and best of luck, -Jake.
Man, I have a power supply here, and instead of a TL431, it have a OP-AMP LM101A. Did you ever found one with this kind of feedback circuit? I'm trying to understand it to hack...
The TL431 is essentially the same. Inside, it is basically just an op amp configured as a comparator with a zener reference. The modification I showed here is how to change the feedback loop of the op amp. Most op amps have trouble handling the same amount of output power as the TL431 but on the equivalent circuit in the TI datasheet for the TL431 it shows an NPN transistor on the output of the (equivalent) op amp. A similar transistor configuration on any other op amp should work the same. That isn't really needed though in this application. Most op amps can power an LED sufficiently. That is all that is happening in this circuit. When the comparator reaches it's threshold voltage it switches on the LED inside the optocoupler. Then that signal is transmitted through the plastic to the phototransistor/reciever on the other half/primary side. Technically this comparator circuit could be done discretely but creating sharp signal transition edges would be more challenging. I looked at the summary specs on TI for the 101a. It looks like that op amp is designed for temperature stability. It also has a larger input voltage range than the 431's. It is unusual though, because the 431's are practically cost free in large quantities for production. The 101a would have cost several cents instead of a fraction of a cent for a 431 not to mention all the space and configuration components. The 431 is even available in a SOP8 package like an op amp if the package space were an issue. Although the most common 431 is probably the SOT23 version which wins the board footprint size/space battle every time. My first thought was that they may have had some kind of dual op amp feedback loop, but after looking up the 101a, it is just a single channel op amp. It's definitely odd, and something I haven't seen before. -Jake
A really simple fundamental explanation of the circuit is that the optocoupler is just a switch. As soon as the controller 'sees' the switch has been flipped on it holds the output voltage at that point. If you think of the moment the switch is flipped like the line on a road, the controller want's to follow that line as closely as possible. It will have a tendency to go barely under that line but is much less likely to go over it. When something changes, like when a load is connected, in a fraction of a second the voltage will drop due to the change in current/resistance. This will turn off the optocoupler switch, and the controller will adjust the duty cycle of the signal on the primary side to compensate until the switch is flipped on again. All that is happening on the secondary side is a circuit that tells the controller "👍 or 👎" There are multiple ways of designing this, but in the end this is a design that must create a yes or no feedback...so long as there is only one feedback channel/optocoupler. If there are two like a TL494 or KA7500 it's a totally different thing.
@@sigmawarrior.fokeryou Yeah it's just an on off switch. Another general rule. Op Amps are designed to create an extremely large amplification factor or "gain" from input to output. This is needed so that they are very predictable with a feedback loop. When a single transistor is used in a circuit as an amplifier it has much lower gain. If you want to build something analogue based on a single transistor like what is inside an optocoupler, you'll need to tune and bias the circuit for every single device made because no two transistors will operate exactly the same. If you look at some devices from 30-40+ years ago you might find a few circuits like this. They will have lots of trim pots, matched pairs of transistors, and other tricks to make complex circuits that just were not available as integrated circuits back then. Nothing is made like this any more. Any design you will find now employes circuits with near infinite amplification and a feedback loop to control the amplification. These types of designs do not rely on properties that are subject to variations between individual components. So what does this mean? Any time your looking at a circuit like this where there is a single transistor and/or LED, it is safe to assume they are operating digitally as an on off switch. There is no way that these could be used reliably to transmit analogue information in a mass produced product. -Jake
Sir this tutorial is extremely informative. But due to my lack of knowledge of SMPS principle. I wanted to ask you, i have a 14 volt 3.6amp psu. I would somehow change the voltage to 5 volts by the way you explained. But is there a way to increase the current output. I waant to convert this 14 volt 3.6 amps psu into 5volt 6amps atleast. I am planning to make a multiport usb charger out of an old psu lying around. If this cannot be done with this 14 volt psu, please suggest me some other ideas or guide me how to achive 5 volts with atleast 6amps Please guide me sir.
@@anirudhkundu3567 There are many many issues. The biggest one is that the original watt rating is the best case scenario for the power output as designed. As soon as you change the output voltage, you also take away the efficiency, and therefore the wattage rating will decrease. SMPS design is all about taking several different variables that are about like dots spread out around a circle and making them cross over at a certain point. There is some margin of error and flexibility available at the crossing point. The points are the switching frequency, controller circuit design, and the transformer properties/windings ratio. Those are just the main issues there are more too. If you change the voltage it will impact several variables. Nothing is free here. If you change one thing a little bit it will affect the other things a little more. You don't need to understand the entire system. It's outputs will operate accordingly. If you want to change the voltage 120% you should expect the current to go the other way by about 150%. By other way, I mean in the negative direction. The way you don't want it to change. I demonstrated how to make a simple "bench power supply." Not how to hack and modify a SMPS for use as that's not even possible. This is intended as a tool that is never left unattended or stressed near it's limits. This is a cheap way to make something similar to a LM317 bench power supply. If you want to build a SMPS for some specific application, you need to take the time to learn and develop a comprehensive understanding of power supply design. I can't help you with that as we would be like two blind people walking around on a mountain. I'm no expert at this stuff. I just shared something simple that I learned here ;) Good luck. -Jake
You can't safely go lower with this kind of design. You need to add another circuit. This is why I uploaded the second power supply video showing how to mod a phone charger circuit. The only way to make a mains power brick go lower is to mod the primary side and that absolutely requires a complete understanding of all of the engineering and math involved. If you were to mod the primary side to go lower and then tried to adjust it for maximum output there is no doubt something WILL get smoked or worse. There are a few ways to make your own SMPS devices that have more adjustment range but they all require more circuitry. I'm building the transistor tester because I need to measure small inductors before I can detail more complex SMPS designs. I have one SMPS project 3/4 complete at over 15 min of fully edited footage and another 2 SMPS projects where I've built a prototype to enable me to design a video around.... More to come ;)
I've built second one, it is adjustable 5v1 - 19v from 19v Laptop charger, but no easily to adjust @16-19v & smooth adjustable on 5v1 - 12v, any suggestion?
Foldback current limiting would be easy to add. Just use a Power NPN transistor on the ground side. Instead of connecting your load from the positive voltage to the ground terminal, you add the Power NPN and a 500ohm potentiometer between them. The ground side of your load connects to the Collector of the NPN. The potentiometer is connected like a rheostat (leave 1 end connection disconnected). Connect the other side of the potentiometer to the Collector of the NPN. Then connect the adjustment wiper terminal to the Base of the NPN transistor. Lastly connect the Emitter of the transistor to the old ground terminal of the power supply. This is the simplest Current Pass circuit. The main drawback is that you lose a diode drop voltage across the transistor. This doesn't matter if you can adjust the voltage of the power supply itself. The diode drop is typically around 0.7 volts. If you use this circuit with a fixed supply like a 5 volt supply, you would only have around 4.3 volts to power the circuit. Like I said, if you can adjust the supply this doesn't matter. The other way to think about this NPN + Potentiometer circuit is to think of it as a Power Potentiometer with a Current Bypass transistor. You could always buy a 100W Power Potentiometer with a resistance in the range you're looking for. These cost a lot. The circuit I described will even work with a trimmer pot (although these are not intended for regular use and adjustment). You can build this with almost any Power NPN the classic 2N3055 works fine. Just make sure you use one with a high enough current and voltage rating. Build a couple of these, they are great adjustable dummy loads for testing lots of stuff. hope that helps :-) -Jake
Upcycle Electronics thanks jake and i know the method with the npn transistor and that’s just waste of power it will be better with an op amp but i was asking if you could make varible constant current inside without any parts outside
It depends on the circuit. The part of the circuit that determines this this the auxiliary power winding on the primary side. When this type of power supply is plugged in, there is a high value resistor that provides just enough current to get the control chip started. Once the chip starts working it draws full power from an auxiliary winding, plus rectifier/diode and a small filter cap. Sometimes this part of the circuit will also have a zener diode to provide additional over voltage protection. The auxiliary winding must be designed so that it produces the right output power when/while the secondary output is running at the desired power. There is usually a lot of room to mess around with this part of the design. These chips usually have a supply voltage range from around 4 to 30 volts. Most of the time the auxiliary winding is designed for the absolute maximum efficiency point of operation which is somewhere around 10-15 volts. This is done because the power lost in this part of the circuit is the primary cause of idle losses when no load is connected. When you lower the secondary output, eventually there will not be enough power coming from this auxiliary winding and that will create the low voltage limit. This circuit is sometimes replaced with an additional power supply. If you understand mains supplies well enough to do a thing like this, the low output limit will just be the control chip's internal reference voltage. The upper limit should be set like I showed in this upload. Don't attempt to exceed the power rating marked on the power supply already. Power (watts) = voltage x current. If you raise the voltage you must lower the current, but the power must remain the same. You should also assume the power output that was originally designed into the circuit is going to be the most efficient point of operation, the further you get from the original output, the further you should de-rate the power. If you want a 20% higher output voltage, assume the totally power will need to decrease by around 20% due to inefficiency. This should keep you from saturating the magnetics, at least in a simple design like I've shown. Still I wouldn't recommend leaving such a thing unattended if you make large changes to the output. I still have and use the one I showed in the beginning. It works well from around 7.5 to 30 volts. I've never left it unattended, but it's handy for quick testing of circuits. Hopefully this helps :-) -Jake
Strefa PC I wouldn't recommend it for electronics. It's an "Ideal" branded meter I got from a big box home improvement store here in the US many years ago. The resolution is terrible. You can get a far better meter for the same $60-$100. This is the kind of meter you use when wiring house Mains or in an industrial setting where you want to measure 240v or 480v. If your looking for a meter for electronics the best value is the Aneng 8009 aka the 9999count meter. There is also another version branded as a bSide Z302 (9999 count) meter for the same price but is produced in smaller quantities and takes a bit more searching to find. If you shop around (AliExpress) you shouldn't pay more than $25 to $30 shipped for this meter. For around $40 there is a new Aneng meter that has a 20k count. It was just released a few months ago and is another remarkable value. If you are bent on a meter you can be confident using with high voltages AND use for electronics, go check out the ones Dave Jones designs and sells through his EEVBlog website. If you want an outstanding bench DMM (not a handheld portable) get on eBay and find a hideously ugly Keithley 197 for less than $100. I paid just under $100 for mine ($80 + shipping). That's a well documented 200,000 count meter. I think I paid around $80 for the Ideal meter ages ago. IIRC it's a 1999 count meter. It's definitely not a good meter for electronics work. I got it long before I got interested in the hobby. -Jake
My list of power supply references I have saved: th-cam.com/play/PLPIwHuVy9EyOtJ_LYw_Rc8oFvBd5TUTYN.html Upload explaining a UC384× circuit: th-cam.com/video/B19rB_FR5Mk/w-d-xo.html Danyk's hand drawn schematic from that video: danyk.cz/reverz44_en.html -Jake
این یک آموزش DIY است که نشان می دهد چگونه من منابع برق ارزان را تغییر دهم. من از 5 مرحله ساده برای پیدا کردن و جایگزینی 2 مقاومت با پتانسیومتر استفاده می کنم. . این برای من کار کرده است من از لپ تاپ و منبع تغذیه چاپگر استفاده می کنم. این کار را انجام ندهید، مگر اینکه خطراتی را که در هنگام استفاده از خازن های ولتاژ بالا و ولتاژهای AC درک می کنید را درک کنید. شما مسئول ایمنی خود هستید. یک اشتباه کوچک می تواند شما را بکشد. اگر خطرات را کاملا درک نکنید، آنها را نگیرید. هنگام تنظیم ولتاژ نهایی حداکثر باید از امتیاز ولتاژ خازن خروجی اطمینان داشته باشید. اصلاحیه ها: من 2 مقاومت از منبع خروجی DC منبع تغذیه را تعویض می کنم. این دو مقاومت یک تقسیم ولتاژ ساده ایجاد می کنند. تقسیم ولتاژ به یک تنظیم کننده شونده zener متصل شده و خروجی آن به یک optocoupler وصل می شود. شما لازم نیست که مدار را بدانید و یا آن را درک کنید. optocoupler همیشه یک تراشه خم شدن 4 پین است که عبور از مانع انزوا را در زیر ترانسفورماتور. معمولا برای تقسیم ولتاژ و برخی دیگر برای مقاومت در برابر ولتاژ و یا پین ولتاژ ورودی رگولاتور شنت، معمولا دو مقاومت وجود دارد. نگران نباشید. خود تنظیم کننده شنت همیشه TL431 است تنظیم کننده شنت یک اتصال دارد که خروجی کامل ولتاژ دارد (یا نزدیک به آن (بی طرفی))، اتصال دوم در زمین است، و آخرین اتصال در 2.4-2.5 ولت (Vref) خواهد بود. فقط به کلاههای خروجی نگاه کنید تا نشانهای مثبت و زمین پیدا کنید. معادله ساده برای خروجی بازخورد است Vout = (1 + R1 / R2) * Vref من نمونه ای با مقادیر این ویدیو نشان داده ام. این معادله به 12.44 ولت رسید. اگر به دقت تماشا کنید، در واقع 13.44 ولت دریافت کردم. این به این دلیل بود که من نمی خواستم همه چیز را با پیچیدگی نشان دهم که چگونه یک تفکیک جزئی کوچک در پین ولتاژ ورودی TL431 وجود دارد. اگر متوجه شدید 12.12 ولت قدرت خروجی به دلیل مقاومت بیشتر از آنچه که من با پتانسیومتر جایگزین کرد، نبود. تنها دلیل ذکر معادله این است که مطمئن شوید که خازن خروجی (ها) را به طور تصادفی منفجر نکنید. دقت معادله مهم نیست، فقط تقریبی است. مقاومت می تواند شناسایی شود، R1 بین ولتاژ خروجی و پین Vref متصل می شود، R2 بین پین Vref و زمین متصل می شود. 2 پتانسیومتر را انتخاب کنید که نزدیک به مقادیر این مقاومت است. مقاومت الكتریك R2 باید یک گلدان حرارتی باشد زیرا شما فقط برای تنظیم حداكثر ولتاژ خروجی به آن نیاز دارید. پس از حداکثر خروجی، از پتانسیومتر شما با R1 جایگزین کنید تا ولتاژ خروجی را تنظیم کنید. محدودیتی برای ولتاژ خروجی بالا و پایین شما با هر طراحی SMPS وجود دارد. این کار با مشخصات ترانسفورماتور، تراشه کنترل کننده فرکانس و زمان خازن انجام می شود. اکثر منابع تغذیه ای که من تغییر دادم از حدود 7.5 تا 22 لیتر ولتاژ پایدار است. عرضه می تواند بالاتر یا پایین تر از حد متعادل باشد، اما هنگامی که یک بار استفاده می شود، آن را می کشد. تنظیم زمان اول: ترفند برای ساختن تنظیم برای اولین بار این است؛ ابتدا، معادله را انجام دهید تا ولتاژ حداکثر را با هر دو پتانسیومتر تعیین شده در مقاومت کامل بدانید. دوم، بررسی کنید که هر دو پتانسیومتر به عنوان رئواستات تنظیم شده اند (تنها 2 از 3 پین به سیم کشی هدایت می شوند)، و آنها را به ارزش مقاومت کامل خود تبدیل می کنند، یک متر و این را بررسی کنید؛ سوم، روشن کردن، و به آرامی تثبیت کننده ترموم (R2 / 5K پودر ترمور در این ویدئو) را در حالی که تماشای یک مولتی متر متصل به خروجی را کاهش می دهد. حداکثر ولتاژ خروجی خود را تنظیم کنید. پس از این تنظیم، هرگز این پتانسیومتر سهبعدی را لمس نکنید. تنوع خروجی شما کاملا توسط پتانسیومتر R1 از این نقطه به جلو کنترل می شود.
Does that help? This video is an example of how I do a modification like this. If my example doesn't make sense visually, you probably shouldn't try anything like this. I've simply tried to show how these power supplies operate just like any other adjustable regulator. There are big limitations and compromises made when adjusting the voltage on a device like this. My intention with this video was to show a different way of thinking about voltage regulators and how they work. I do not want to create a schematic or the detailed information. The person that does not understand the fundamental information I demonstrated here should avoid a project like this. These devices are very dangerous.
User 820 There is a simple and complicated answer. The simplest answer is this example: lh3.googleusercontent.com/-AJRuQwP_BQg/WZB87RRtAxI/AAAAAAAAAaQ/KsJZt0lEyCkKjZMXmY4RoTXv9-teNUKQQCHMYCw/s1600/Screenshot_2017-08-13-09-20-04-1.png I am comfortable using the current rating of my supplies at the original voltage rating. I have used about 70-80% of the rated max current of a supply within 5-8 volts of the original voltage of the supply. Once the output voltage gets farther from the original design you have to delve deeper into the design to find it's limits. The timing capacitor and transformer design need dramatic changes in order to create several amps of current reliably. My supplies that I have modified have no problem supplying 1-2 amps of current at higher and lower values, but I have not pushed them to their continuous voltage/current limits. I have tested brief outputs with low value resistors and seen 3.4A at 22V from the supply I showed modded in it's original enclosure. The frequency and duty cycle of the output transistors could become a major issue if you push them continuously outside of their original design parameters. The values used for the resistor divider have no direct effect on the output current. They can be anything from 1k to about 100k. It's the ratio relationship that is important. I used 5k and 20k because they were close to the original resistor values and it helped simplify this example. They could have been 2k/5k, 2.5k/10k, 20k/50k, or any other combinations that can achieve a similar 1:4 ratio. All of the controllers I have shown, and indeed most cheap SMPS controllers used in laptops/printers are called "peak current controllers." This type of controller varies the frequency of the AC signal to change the output voltage. I can give you a lot more info if you really want to know the grity details, just ask. Fortunately, the method I showed here only biased the secondary side of an isolated SMPS. There is another bias point on the other side of the optocoupler (the 4 pin IC). This limits the potential for someone to really do damage or blow up a circuit. If you have a specific project you want to power at higher currents it's best to start with a SMPS at close to your desired output. For general bench projects my supply has yet to reach its limits. I will be showing another MC34063 example soon (current project I'm working on) that is part of an AC to DC design anyone can build. I'm trying to break down the details even more and show some of the limitations and functionality of a simple adjustable SMPS. I will try and demonstrate the characteristics on my oscilloscope once I finish the project enclosure build. I plan to demonstrate the design from this video, the current project, and compare it to a cheap commercial LM317 supply I have. That should help clarify the subject a bit.... ...and this is why I don't speak in my videos to date.... Describing electronics is a nightmare.... All of that said, I've read a ton, and experimented a lot, but I'm not an electrical engineer. I make mistakes. I am not an expert. All I can tell you is this is what I've experienced and seen for myself. I have seen repeatable results and shared them as such, but I can't represent them as anything more :)
genuino 301998 It depends on the controller and how it is configured. Usually you can't get below 7-8 volts on a device originally designed for 12-18 volts. There are several factors that cause this. It's mostly safety related factors built into the design to prevent catastrophic failures.
YouTubiando The maximum power output should be whatever the unit was originally rated for at the rated output voltage/current. As you get further from this number the efficiency will decrease and you will get less total output.
First of all, congrats on the unique idea and the fantastic job and thanks for taking the pain to share it. My apologies if you are vocally challenged, but if not, by going mute you have unnecessarily complicated a subject that is anyways not easy for a non-electronics person to understand. I completely lost you in the part of the video that has the number 1-5 annotations. The great pain you took to write it all down in the description section does not help at all because it is not the same as a verbal commentary or text annotations explaining the steps as these are being performed.
percuma adapter seperti itu dibuat variable, kalau cuma sekedar merubah besaran output bukan hal yang sulit saat diberikan beban protek aktif sehingga adapter itu tidak bisa jadi serba guna so for what. cobalah buat konten yang merubah adapter menjadi high load atau menghilangkan auto protection supaya bisa sesuai judul high current.
Ah... no i realized what it was... i figured there was an "on" switch with the potentiometer, and there is not. It's just on if it's plugged in, which the multimeter correctly recorded. It goes from 7.5 volts to 27 volts unloaded... s'alright.
Si hablará y explicará sería bueno ... pero no somos sordomudos ... y estoy más que seguro que si un sordomudo hace este video ..se le enfocaria el cerebro a poner un traductor con señas ...
Sigh. these "I insist on no talking to the point where I slowly make note with pens IN A VIDEO?" are so tiring as a format. Seems like if you aren't going to say anything extra just provide a set of notes or a generalised schematic. You seem nice, knowledgeable and everything, but this format is annoying.
Why is it now nobody likes to speak on there own video. Before the only people who didn’t speak were the one who didn’t want to get in trouble. But these guys are on the up n up I don’t like when they don’t explain what there doing maybe it’s because I’m slow who knows in any case i don’t like it but good videos but can be so much better
Great video. Saw your comment on Electronoobs' video on the DIY BTPS for under $10. Please consider an update to this video with some narration giving more detail. There are too many parts of this video that weren't obvious to someone just getting into this field. For example, some of the test you did with the multimeter and the modifications you made. There are many who'd be grateful for the effort and know what it takes to create content. Cheers
Hey,
So this video is intended for a few different things.
Firstly, all of my uploads have been an exploration of how the YT/Google search engine and suggested content systems work. All of the uploads have titles and descriptions in a number of different languages. The content is entirely void of spoken language for this reason. This stems from my previous ecommerce businesses and an in depth analysis of different sales platforms I've worked with. That was more of a personal curiosity.
Secondly, the entirely intuitive based approach to this specific upload naturally limits the type of person that will find this useful and attempt it in practice. If someone needs a step by step and detailed walk through of this, I don't trust them to be capable of handling the dangers involved here. This kind of thing requires the person to have an intuitive sense of what is happening and the extreme dangers involved. This is very serious, and not something everyone should try. This stereotype probably does not encompass you, but it is reflective of my choice on the content I want to make.
Thirdly, this is an oversimplified example of the similarities between a switching power supply and a linear regulator. People sometimes mention this when talking about SMPS theory, but I've never seen someone demonstrate it in practice like I did here.
I understand far more about this subject now than I did when I made this. It's still a useful example to demonstrate the intended ideas, but it's not a super useful device in the first place.
If you are just starting out, buy yourself a decent power supply for $50-$100 and go from there. This may seem like a big deal or too much to spend, but it is a savings in the long run. I'm literally working on my 6th (bad) bench supply design right now. I've learned a lot, but I can honestly say I would have been better off and would have saved more money if I had just bought a decent supply or two second hand on eBay.
Alan over on his channel W2AEW has shown a couple of older HP supplies he likes to use and are available relatively cheaply second hand. I think I even saved the reference in my Fundamental Power Supply playlist.
BTW, check my playlists if you haven't already. I don't use them as self promotion. I couldn't care less about self promo. Those are all of the references I save for myself and cover a massive amount of stuff I've come across and watched over the last few years. I've been trying to add info in the playlist descriptions about other related playlists. Everyone else can't see my alphabetical indexing system. At least, if you know it's there you might find what you're looking for. I need to subdivide the power supply playlist but haven't yet. It's got more than 100 references. I'm always adding stuff I come across. Playlists like this allow you to find much better references that the YT search system will not show you or feature. YT prioritizes monetized and recent content far more than relevant references. This stems from pop culture stupidity being the primary driving force for the platform. I made my references public because they can be useful for anyone following the same learning curve as me.
Lastly, if you're looking for linear power supply designs, I've been looking at ones based on an old LM723 or CA723. Those are extremely precise regulators. They require a current pass transistor and a little more circuitry than a lm317 but I think they make a decent bench supply. You can find many circuit designs by searching images on duckduckgo or other engines for things like "LM723 power supply schematic." You can always just hit me up for specifics or a reference. I will share anything I know or where I go to find out more :-)
Sincerely,
-Jake
Wow, a lot to unpack in your reply. I read every word and totally get and appreciate where you're coming from. As for danger, we may be increasing it by not offering more information for those who will take risks regardless. On the other hand, we can't be held accountable for what others do with content of this nature. Ultimately, it's up to each content creator to decide what to offer. but you know all this. Cheers brother
Sir,
I appreciate the comments you made about not wanting punter noobs who don't know which end of the soldering iron gets hot to mess with dangerous voltages, etc. I totally agree, but with one caveat:
I totally agree with @cri8tor about the lack of information.
I, myself, have been working with electronics since the 1960's, (when television flyback transformers generated 30KV+ and could knock you across a room!), and have a considerable body of experience, but I still found your video virtually incomprehensible. In my humble opinion an obscure and incomprehensible video is worse than none at all since people will try things, hoping to duplicate what you've done, and end up doing serious damage.
I would really like to try to duplicate your results, but the lack of information is scary.
Do you have other videos like this one - or better yet a web page that provides more detail - so that those of us who cannot read minds, ( :wink: ) can follow along?
Thanks!
I found the description 2 be a solid counterpart 2 the vid itself.. Together they make the entire process Very clear, imo...
It works by modifying the voltage divider for the TL431 reference IC (voltage feedback) or something similar
So basically the IC needs a 2.5V coming from the voltage divider, to trigger the voltage feedback pin
So if you swap out the factory made voltage divider with a potentiometer, you can adjust the middle pin voltage, which goes to the TL431 reference IC
The lowest voltage possible would be 3V, while the highest voltage depends on your SMPS circuit
There is one other factor. With the way these circuits are designed, the control chip must be powered by an auxiliary circuit on the primary side. There are a few ways this is done in practice, but most of the time there is a small set of extra windings connected on the primary side. This winding will usually have a single slow diode like a 1N4007 and a small capacitor of around 100uF@50V-ish. The quality of this circuit is largely determined by one simple thing, if there is a large-ish zener diode in this circuit. These chips usually have an input power voltage range of around 4.5 to 30 volts. This is the actual limiting factor when adjusting the output of a supply. It's this voltage range in combination with how the transformer was wound. Specifically, the turns ratio of all the windings is critical.
When designing a switching converter circuit there are several factors that need to converge around the desired output voltage and current. These factors are permanently baked into the circuit. The windings' impedance, and the turns ratio of the Primary/Auxiliary/Secondary, and the switching frequency are all part of the design. They are not precision type elements though. Picture them like a wide bell curve. The output voltage, current, efficiency, and noise are optimized to somewhere near the top of the bell curve, but there is a lot of wiggle room on both sides of the curve. You can't make the peak or shape of the curve any different, but you can alter the parameters and have the output cross somewhere else on the bell.
All of that said, when the voltage of the secondary side is changed, the voltage will also change on the auxiliary winding. A properly made power supply will also have a zener voltage clamp on the auxiliary power circuit. In most, if not all, countries a mains voltage switching supply must be able to have a failure on the secondary feedback circuit and not catch fire or go critical. Think about this for a second. If the feedback loop fails, that means the switching circuit on the primary side is going to push out the maximum dutycycle non-stop. So a supply must be able to survive this condition.
The reverse aspect is the minimum voltage that the control chip can operate at. The chip doesn't require much current to work. This means that it's auxiliary winding can be quite small and insignificant. However, if the voltage is changed it will vary quite a bit. The single diode half wave rectifier doesn't help either. The way the Auxiliary winding is setup is the primary output voltage determination factor. In principal it should be wound to the middle of the chip's input voltage range. This will be a voltage close to the chip's lowest quiescent current. The Q current is a major factor in the no-load efficiency. Some people are cheap though and will split a thousandth of a penny making an aux winding that is as few turns as possible. If the aux winding has too low of a voltage the chip will have an under voltage shutdown function. This will make the output circuit cut out off and on. Most supplies can't go all the way down to 2.5v unless they were designed for close to it in the first place. If you know what you're doing and power the auxiliary circuit separately, then yeah, it will go down to 2.5 or below depending on the feedback resistor setup/mods. It will be noisy as hell though. It was designed for a pwm dutycycle optimized for the output, so the lower voltage makes the thing come on, go off, and take a nap while the output caps have an aneurism.
Basically, it's not good to use this type of mod at the extremes, but if you have say a 14 volt supply at 100 watts and need a 12 volt supply at 80 watts this should work fine. It's not much different than a linear regulator. Don't hook up a 3.3 volt reg to a 12 volt input and expect to get the datasheet's max current figure. The magic smoke will find you. While there are more factors with smps design, it's still the same basic thing.
-Jake
@@UpcycleElectronics Yes, I know, but it's only used as a voltage sensor, so most of the current doesn't flow through it.
When the voltage divider middle pin reaches 2.5V, the circuit only sends a HIGH signal to the chip, which regulates the duty cycle accordingly.
And the current through the 431 is very low, only a few mA.
Also PC ATX power supplies that have KA7500 have a built in voltage reference, you just hook up a voltage divider (potentiometer) between the power supply output, and when you twist the potentiometer, the voltage on the middle pin changes, so the IC will always try to see 2.5V at its feedback pin, and so the output voltage changes.
Also my power supply will rarely get current times when the load goes above 1A, and since my SMPS is rated 2A, it should be safe.
I'm using this modified unit for my mini PCB drill, the motor current is ~ 250mA, occasionally it exceeds 1A, when I'm drilling something, but these currents are for short periods of time.
@@adamsucksatyt
NO! A TL494/KA7500 is entirely different. It has 2 feedback loops. Both must be setup in tandem. Don't try to adjust a dual feedback chip. That's a whole different can of worms.
Chips like the UC3842 have a much more simple current limiting circuit. The dual feedback chips must have a calculated design. I have several power bricks that were commercially sold as adjustable. A couple are just like what was in this video. Others are a UC3842 like circuit with an MC34063 on the secondary side with an adjustment divider in their feedback loop. I've never seen someone try to adjust a dual feedback loop controller. You don't want to mess with something like that either. Those multiple voltage rail supplies only regulate 1 rail's voltage and current. That primary rail must have the majority of the load on it or the supply will fail in short order. Leaving the other rails floating can cause problems too. Those are purpose built tools. They are not really flexible things to mess with... I was told this, and it still didn't stop me... so whatever :-)
...just saying...
I built a box with 5 modified power bricks inside as a basic supply. If you know they are all isolated from mains, the outputs can be placed in series or parallel just like batteries. That's a much more flexible type of setup. Just be absolutely sure the supplies are totally isolated and the secondaries are floating.
@@UpcycleElectronics Dude calm down a bit, I don't know everything
What do you think, does a 15 year old know all IC datasheets and make right assumptions?
Also I'm talking about TL431, not TL494
@@adamsucksatyt
ATX computer power supplies are almost always TL494 based.
If you haven't found the channel Diodes Gone Wild go check out his stuff. He reverse engineers a ton of these types of supplies.
I'm just learning this stuff too... although I have a few years head start. I've been saving references for awhile. They are public too. These are 2 playlists from the 6 I saved about power supplies.
th-cam.com/play/PLPIwHuVy9EyPLzeyuEqKVK_LmWVQNzkZ3.html
th-cam.com/play/PLPIwHuVy9EyPppZvFMXCheasl4DwlS2r6.html
Thankyou for the clear vid and how to do. I thought u did a great job without saying a word, was very clear to me anyway.
So i collected lots of diff p supplies and had a go - i got 2 out of about 6 working. Also i found most didnt have just a simple divider to the TL. They were 2 resistors parallel and one series, combos like that. No matter which values or what combinations i removed and replaced with equiv value VR, i could never even adjust the output by more than a volt. However, with the couple i did get to work, like he stressed in the comments, the output voltage was unstable. Like, REALLY unstable the further down it goes from its designed voltage. I found once it got down near even half, it fluctuated up and down on my multimeter without even pausing, and that was without any load, ie, the output wasnt connected to anything at all, just multimeter measuring it. So i concluded, the ones i tried anyway, and with my very limited knowledge and skill, that doing this was not going to result in a new home made adjustable voltage power supply. Not something actually useful and reliable anyway, just limited use in its upper range. Near to what it was designed for. It was fun tho, and now im much more familiar with the little opto feedback to tl voltage ref section of these supplies.
The supply that i had many of and thought would be the most useful when modded, i couldnt do at all. 24v, 10A, brand new boxed still, tl491, had as upper and lower resistive dividers something huge, like megohm or something, and the other so small my meter couldnt see it, like 1R or sub ohm, i cant really remember. I tried anyway, with closest VR's / pots i had, but no luck.
You had me impressed at first glance. Around 1:18 I felt your pain. If you use a very rounded chisel and a hammer you can easily remove the cover to the point that it can be easily glued back together. I'm very ocd and had to find a way lol
Rafael Roman
A hammer and chisel could work.
In a past life (about a decade ago) I owned a small autobody shop. I specialized in plastic and small panel repairs. Most plastic automotive panels are a similar type of ABS plastic as power supply enclosures. One of my specialties was plastic fusing and plastic reconstruction on rare panels and bumpers that are not reproduced by 3rd party manufacturers. This type of job uses a large variety of methods to reconstruct a surface. The most common and effective way to repair cracks or 2 ABS plastic surfaces you want to connect is with a soldering iron and some fine wire mesh....
..Anyways the "just use an iron and melt it" idea is the same. These power supply enclosures are designed with a interlocking lip, 'tung and groove' style. The plastic is welded/melted together using an ultrasonic welding machine (welded using ultrasonic friction). This production process seems to be kept on the ragged edge minimum for the amount of effort needed to fuse the two halves of the enclosure. I have seen several that easily come apart with a few minutes of light persuasion.
The trick is to be patient with channel locks or a soft jaw vice. Only apply enough pressure to flex the surface on one side of the enclosure while gradually distributing the pressure all the way around the enclosure. It will usually take 5 to 10 minutes of time to completely work the thing open. If you do it this way you will get a nice little project enclosure for free, about 9 out of 10 times. The edges will have some small imperfections, but they are not very noticeable if your patient while opening the thing.
When I reuse one of these little enclosures I use one of 3 options. One option is to use a soldering iron to embed a nut in the plastic on one side, adding enough extra ABS plastic from something else I have around to thoroughly bury the nut. Then I drill a hole through the other side and bolt it together. Another option is to use hot glue to join the two halves as this is fairly easy to reverse if you really need to. The last option is to use a soldering iron to reweld the two halves together by remelting the flange. This is a permanent fix though. When you melt ABS plastic with a small heat source like a soldering iron, it tends to get more brittle and it loses some of its functional properties. Once this is done it gets harder and harder to rework. I learned the hard way a long time ago, you only get one chance to rework and get it right.
Just a bit of my experience from ages ago.
....Alternatively, I recommend not being cheap like me and buying enclosures for a few bucks instead of wasting your time hording such useless things. I have 13 of them in my drawer right now ;/
The 'Holy Grail' of power supplies are the ones that bolt together from the manufacturer. They usually have interesting components inside as well. I even have a couple of wall warts that are held together with screws. The other 'extremely (functionaly) valuable' SMPS power bricks are the HP type with the 3 terminal plug on the DC side. Most HP supplies have the 3 terminal plug, but only about 1/3rd actually use it. The label will tell you, or if you look at the plug end just look to see if it has all 3 pins populated inside the plug end. This type of supply is dual polarity. They achieve this simply by using multiple secondary windings on the transformer. So the feedback circuit is just like the one I showed here. I tried to build one a few weeks ago to see if it is the same and it works fine. I can adjust that little supply from 12 volts to 20 volts with a positive AND completely isolated negative polarity. I can also wire them in series for 24volts to 40 volts. The circuit is earth mains referenced from HP so you must disconnect this wire, (actually a whole copper PCB plane is used) in order to isolate the power supply circuit.
The one I modded is rated at +32volts at 1100mA and +16volts at 1600mA on the label. I think they must show a positive voltage on both channels on the label...cause people are stupid... but yeah, it's +/-16volts from the factory. I've tried to mod some other dual polarity SMPS designs like the old Dell's but they are super complicated PWM units that have multiple feedback loops.
-All the best, -Jake ;)
Upcycle Electronics thanks for taking the time for that extensive reply. I have a box of 12v 5Achargers and I mean I have almost 100 so when I saw this video I had to see how it was done :)
I wonder if you where able to pump 20v on only 1.6A supply, could I get a hover voltage with stable current with a bigger supply?
Rafael Roman
I don't get out much after a broken neck 3.5 years ago. I tend to be a bit too chatty at times ;)
Inductive loads are tricky. About half of the supplies I've tried get motors moving. Usually a large Capacitor can help get them started. A vacuum might be a challenge. Usually the generic supplies seem to handle motors and such a little better. I use one branded "LITEON" 12v 2.67A to run an old skool Milwaukee hand drill (1st gen cordless) and a small dust buster Black and Decker hand vac. It works as well as the batteries if not better... and it's unmodified. I've tried others rated at 2-3 times that many amps with no luck. I've even tried using a NTC to provide the start-up resistance, but even that failed. I think the load trips some part of the protection circuit on some chips, but I'm not positive. I should probably take apart the ones that work and some that don't and put them on a scope to figure it out, but I'm slow to do things on my work bench especially lately. Maybe I'll get around to that one of these days.
Upcycle Electronics I just subscribed to keep up with you buddy. Cheers from Puerto Rico
Upcycle Electronics,
You surprised me ! Thanks for efforts
Hey man!! Firstly just wanted to say thank you for making this content, and especially then spending so much time in the comments helping/answering questions.
I wanted to make the same complaint that a load of people already did haha (please explain more etc.) but I actually read far enough in the comments to realise why you made the video as you did and I appreciate the logic. (You could pin the comment which specifically mentions this, as a lot of people won't get that far down before they ask you the same thing!)
So anyway I'm just at the beginning of this learning journey and yes this is quite a lot too much for me to understand at the moment, no problems I can deal with my frustrations for now, take a step back and I'm sure that in some weeks or months or even years haha I'll come back to this and find it more fascinating, accesible and useful!!
I did open a laptop supply I have which malfunctioned a few years ago, fully aware that I don't understand it very well and that it's dangerous. I looked at the voltages on the multimeter across the various capacitors and watched them slowly going down which was cool (so I guess it has a bleed resistor for safety as it should right??) I then put a load across and watched them empty, and something I didn't expect or understand happened which is that after they drained, they slowly started regaining voltage a couple of mA a second. So today I will look up this phenomenon and try to learn about it. I also learned why the power supply failed, I think, which is that two of the resistors are a long way from their stated values - something else to go and look up (the list is looong)!!
And what I have learned from here is that I think I will buy a cheapish lab power supply and make a DIY one just for a project, but a little further down the line, when I can understand it better.
So again thanks for your help, I have subscribed to your channel and hope to be able to make sense of such videos in the future, and feel confident about safety and making useful and effecttive mods myself one day.
Thanks for reading a little deeper into the comments. I pinned one of the ones where I replied with my reasons for not talking in uploads.
The capacitor reacquiring charge is a common thing. I've read about it, before but don't recall what it's called from memory right now.
I'm currently on a tangent learning about digital electronics, linux, networking, and more specifically trying to wrap my head around how a UEFI bootloader system works. So, analog electronics are on the mental back burner.
You also mentioned measuring resistors. Be very aware that it is virtually impossible to measure resistors and capacitors when they are in circuit. The device making the measurements is using ohms law. This means it is applying small voltages and/or currents to detect changes between the probes. As a general rule of thumb, unpowered circuit boards are ravenous beasts willing to devour every spare electron they can acquire. When testing a component in circuit, you need to have known conditions, if you want anything accurate or even relevant. This is why, in the upload, IIRC, I either removed and measured the resistors or I used the numbers printed on them to state the values. I used continuity in circuit just to figure out what was connected to where. I had to measure the voltage divider while under power in order to figure out the power, ground, and voltage reference node. I don't think I ever showed a resistance measurement in circuit. If I did, I apologize, as that would be pointless.
There are ways of measuring some of these components in circuit. The tool used is called a LCR. It measures using frequency instead of using simple voltage/current relationships. The LCR is actually measuring impedance (resistance@frequency of X // aka AC Resistance). Don't worry about that too much though. Regular ohms law type stuff is much more important to grasp and use.
The LCR stuff is really only useful/worth the investment for people that understand and can manually design radio circuits and frequency filters from scratch. Most of us will never get very deep into that can of worms. You could spend a lifetime going down that rabbit hole and still not fully understand the subject. In other words, it's not a path you should wander down randomly. That should be a well planned trip with lots and lots of planning in advance.
Most people will find and use premade circuit elements when they need a filter circuit. You'll wind up buying components with specific values, and from reputable sources so that you don't need the expensive tools to make the LCR measurements yourself.
Lastly, resistors don't usually change values. When they fail it's usually something else that caused a lot of current to move across them and they burn. You'll KNOW when they fail. It's that big charred black mess.
Usually, switching supplies fail due to capacitors or the switching transistor. This is complicated. If there is not a snubber network on the mains side, there will be a voltage spike on the main switching transistor that can kill it with normal use. Transient voltage spikes on the mains power line in your area are also common (everywhere). MOV's protect against this, but most cheap PSU's do not have them. The mains rectifier diodes are also a common failure mode. Bad caps are the most common failure mode though. When they fail it usually kills the controller chip.
Get The Art of Electronics 3rd Edition. Search for the PDF online. Uni students pirate scan and post them to combat ridiculous book prices. See if it's a good fit for you before buying it (yes you need the 3rd edition). Also get a copy of Hacking the Xbox by "Bunnie." Second-hand copies are dirt cheap on eBay. First, it isn't about the Xbox. Second, your interest level in the xbox/gaming is irrelevant to the value of the book. Those are my 2 most useful "general references." I still don't have a copy of the ARRL, but that's another good one. I've gotten library copies multiple times.
Hopefully this helps :-)
-Jake
@@UpcycleElectronics Hey Jake thank you so much for taking the time to reply and the suggestions and information.
Funnily enough I've used linux for years and I remember trying to get my head around UEFI boot systems a few years ago - I don't think I got very far with the understanding part haha but enough to do what I needed.
Well as with everything, the more you know the more you realize you don´t know, and evidently I'm right at the beginning of this journey. Feeling very motivated to learn more though! I managed to do a super easy 3d design on Tinkercad (kid's online CAD programme, it's awesome if you haven't heard of it), then I made all the pieces out of "plyboard" - using PVA and cardboard in a 3 way ply is suprisingly strong - and built the box. I rescued a broken 3w bluetooth speaker which sounded awful before, and managed to combine it with the solar panel and [6x] bigger battery from a solar charger which all just happened to be the same voltage/chemistry....and it sounds amazing and works really well. So you can imagine I got really enthusiastic, and have now taken apart a few broken bits and bobs and have a few projects in mind.
Inevitably I am now realising that to do anything mildly custom, which isn't just connecting up pre-made modules, there is actually a huge basis of knowledge which I need to chip away at if I want to slowly build a decent understanding.
Yesterday I tried to measure the amperage, in the same way as one does the voltage, and couldn't understand why I shorted the circuit lol. Then of course I thought about it a lot, and went and watched a simple tutorial, and everything became so obvious. Fortunately all low amp/voltage things so no damage to the multimeter, and after the obvious homer simpson "doh" moment and feeling stupid, it actually felt great to make sense of the obvious.
Today you've given me another one of those moments. I was kind of thinking about it yesterday as I was measuring the resistors, but obviously didn't make the click.
Today I bought a DIY kit for a bench power supply, and I think it might end up sat on my shelf for a fair while while I get constructing some super simple circuits and do some more studying, but it's a nice goal to aim towards and would be fundamental for a couple of things I'd like to do.
I have a copy of "Make: Electronics" which I am going to get started on, and I will check out the other books you recommended too.
Good luck with your various endeavours and if I get a channel up at some point to have a look at any of these projects (I think the speaker idea for example, would actually be a cool learning project for other people too) I'll let you know!!
Alternatively if I never comment again, we can assume that I didn't fully understand something about high volts at lots of amps and touching bare wires or something like that lol
Thanks again man!!
@@TheKoodus
Hey, if you ever get around to making some content, I am your first subscriber :-)
Electronics is a deep hobby. If you go down this rabbit hole you may find it's a lifer.
I started into the hobby after a car hit me bad, like really bad, on a bicycle commute to work Feb 26th 2014. No sympathy needed. Just saying, I started into hobby electronics thinking it'd help me burn a few months to a year while I recovered from a broken neck. I never fully recovered physically, so I've been at this every day since and still feel like I know close to nothing.
Best of luck :-)
-Jake
Brilliant, I have a couple of unused laptop adapters laying around and was thinking of disposing them, but you've just given me an idea on how to re-purpose it. You just earned a sub :D
rene kenshin
Thanks. This is a device I use all the time now ;)
Awesome, I'm a hobbyist myself and this may come in handy :D
Nice work, can you make it with ATX because of it still a big secret?
No. An ATX power supply uses a different type of feedback loop to monitor the voltage and current actively on the main secondary rail. The circuits shown here use a simple resistor to sense the current on the primary side while directly sensing the output voltage of the secondary. This single feedback loop is much easier to modify. If you modify an ATX supply like this it will fail very quickly. That type of design must be engineered for a specific application. There are channels that go into designing circuits like an ATX power supply. Mr Johnnny and Diodes Gone Wild are both good channels for this type of content if your looking to build something powerful. The ATX circuits are usually based on a TL494 IC. That IC is also called a KA7500 that is basically the same thing and is pin compatible. If you haven't designed SMPS circuits before, these designs are very complicated to do. I would start with something like a UC3842 IC design if you really want to build a mains powered SMPS. If you've never built a SMPS before start with the MC34063 buck converter style design. That's the IC used in most older phone chargers for cars (your car is 12v and your phone is 5v so these are simple buck converters). The MC34063 is only capable of supplying around 500mA but it will start you on the path to SMPS design.
Danyk is the content creator for the channel Diodes Gone Wild, if you search the internet for "circuit" or "schematic" and "Danyk" you will find his tutorials about UC3842 SMPS designs. That's probably the easiest place to start.
-Jake
thank you very much , I,m new in electronics but I read more, and yes I subscribed all these channels you mentioned I want to re-engineer control chip in Atx modifying TL494 and protection lm393 with 30v and 5A, can I use Reverse engineering by clone PCB and convert it to Gerber then get the schematic (the secret) ?
and by the way, I love what he did:
www.instructables.com/id/Adjustable-High-Current-Lab-SMPS-Out-of-a-Standard/
best wishes : )
Thank you very much. Great job. I want to know the soldering iron transistor 220V-80W. The old one was burned and I didn't know its number
I like your style of video making. Kudos!
¥
Hello, nice video ! and good explanations for the maths !! can you explain how to act on the primary side to increase the output current ? thanks
The key is to understand that Ohms law always applies. Watts = Volts * Amps. The current/watts rating is built into every part of the circuit. The size of the magnetics, the gap between the two halves, the size of the copper wire used, and the number of turns, are the biggest factors. The diodes, and switching transistors are also selected for a specific output current. When you change the output voltage, you're detuning the circuit. In other words, you should derate the output. If you change the output voltage by 20%, you shouldn't expect more than 80% at the output.
For example, let's say the supply is rated for 10 volts and 50 watts. If you change it to 12 volts, at best you still have a 50 watt supply. You've changed the specs by 20%, so the sensible thing to do is assume a 20% reduction in the maximum output. This means your 10 volt 50 watt supply was capable of 5 Amps max. At 12 volts 50 watts is 4.1 Amps. If you are sensible and derate the power supply by the rate of change, you wouldn't use it for more than 3.3 amps.
I didn't show "how to hack a switching power supply." I showed how to make a little adjustable bench supply. A bench supply is something that is always attended and generally gets light use.
If you have a purpose in mind, than this is not a good project for you. When designing a switching power supply there are around a dozen factors that all come together at one point, the desired voltage/current output at a specific switching frequency. This is like the peak at the top of a hill. There's lots of slope on both sides of the hill, but you aren't going to make a bigger hill.
@@UpcycleElectronics Hello dear youtube expert ! thanks for your detailed answer, i got what you meant by maximum, but i thought that through some fans and increased capacitor, i could increase the output current by 30%.
On another hand, why did you put TWO potentiometers ? i could thanks to you better (not fully yet) understand the voltage divider thing, but why making two variable resistors since the max stays the same : 20Vdc.
@@hichamtassi1753
If you watch the video carefully you will see what is happening with the 2 potentiometers. One is a trimmer only. Once it is set you never touch it again. The second pot is the adjustment setting you use all the time. If you watch carefully I showed how the pots were adjusted. I then connected the power and set the maximum voltage level using the trimmer pot.
@@UpcycleElectronics Ok, perfect then, i tried to change the current through the PIN3 of the SMPS chip, i could only increase the output by 0.5A and it was not linear, so not worth it...
Awesome project I'd love to see a server psu or pair most important a way to adjust voltage and current. Needing a constant current source
Multiple voltage tap types of supplies are usually much more complicated than what I showed here. These simple smps supplies only have 1 feedback loop. This makes them much easier to modify. If you're looking for a good source of circuits like a current source get the 3rd revision of 'The Art of Electronics.' If you look around there is usually a PDF floating around the web for "free."
I save all the electronics references I come across in publicly listed playlists. This is my list for current sources and stuff:
th-cam.com/play/PLPIwHuVy9EyOm6MjX1_J_YW9QKfQcI2Vg.html
And this is my playlist for power supplies in general:
th-cam.com/play/PLPIwHuVy9EyOtJ_LYw_Rc8oFvBd5TUTYN.html
hi..nice job
I have a dell laptop adapter 19.5 volt and 3.35A.
I want to get 12 volt and 3+A from this adapter can you please guide me I'm not expert in electronics please kindly guide simply if you can..thanks in advance
This example is for someone that is just making a little lab bench power supply. The video illustrates how I modify these kinds of power adapters. If you need a specific output for a project or device you need to purchase the appropriate product designed for the application. Never modify a charger's output like this and leave it unattended unless you know all of the math involved to calculate all of the potential failures in the power supply. If you don't understand what I have shown here, messing with power supplies like this is a very bad idea. Switch mode power supplies are designed for a specific output, there is a certain window of adjustment that will still result in an output voltage but it is compromising the design when a different output voltage is set. The modified output is barely useful for the most basic types of bench testing and experiments, but it is in no way a suitable replacement for a properly designed power supply at a specific power output.
What happens if i dont remove the resistance r2 and only add potentiometer to r1?
Thanks jake
Nice video, thank you, keep it up :)
The way you Opening adapter cover is wrong could be damage the rite way is just apply karocine oil aur sprite/ thinner arround the joint then tap gently with screw driver base all 4 side it will open the cover without breaking.
Great,... Is it work?
it is confused, make a note to every step that what you doing, it will be help for us .. anw this is first tutorial that I've seen increasing output voltage - thumbs up guys
TheJavaSync
I tried to explain it in more detail in the video comments description.
I am trying to create content that is as visually intuitive as possible. The title and basic comments are translated into several languages using Google Translate. I'm trying my best to make something that is useful to people who may or may not speak English.
TheJavaSync
I just added to and modified the comments description to add the following:
"I replace 2 resistors from the DC output side of the power supply. These 2 resistors create a simple voltage divider. The voltage divider is connected to a zener shunt regulator, and it's output goes to an optocoupler. You don't have to know or understand the circuit. The optocoupler is always a 4 pin dip chip that crosses the isolation barrier under the transformer. Usually there are two resistors for the voltage divider and some others for biasing the opto and/or the shunt regulator input voltage pin, don't worry about these. The shunt regulator itself is always a TL431
The shunt regulator will have one connection that has the full voltage output (or close to it(biased)), a second connection will be at ground, and the last connection will be at 2.4-2.5 volts(Vref). Just look at the output caps to find positive and ground traces.
The simple equation for the feedback output is
Vout= (1+R1/R2)*Vref
I've shown an example with the values from this video. The equation came out to 12.44 volts. If you watch carefully I actually got 13.44 volts. This was because I didn't want to complicate things by showing how there was a small additional bias on the voltage input pin of the TL431. If you notice it wasn't at the full 12.12 volts of output power because of the extra resistors below the ones that I replaced with potentiometers. The only reason I mention the equation is to insure you don't blow up the output capacitor(s) accidentally. The accuracy of the equation doesn't matter, it's just an approximation.
The resistors are easy to identify, R1 will be connected between the output voltage and Vref pins, R2 will be connected between the Vref pin and ground. Pick 2 potentiometers that are close to the values of these resistors. Resistor R2 should be a trimmer pot because you only need it for setting up the maximum output voltage. After you've set the maximum output, you use the potentiometer you replaced R1 with to adjust the output voltage.
There is a limit to how high and low of an output voltage you can achieve with each SMPS design. This has to do with the transformer specs, frequency controller chip, and timing capacitor.
Most power supplies that I've modified are stable from about 7.5 to 22ish volts. The supply can be biased higher or lower but when a load is applied it will get pulled down."
alright, thank you very much (y)
awesome, waiting for another hacked electronic videos
TheJavaSync
I'm 4 weeks into the current project but I'm a slow partially disabled gimp. I may do another short side video on enclosure building/plastic welding though.
The Next Project:
I am waiting on a transistor tester someone ordered for me. I should get it in the next couple of days. I really want to demonstrate the difference between switch mode and linear regulators. I am trying to show how you can build an adjustable AC to DC power supply using an upcycled phone type car charger and a salvaged transformer. I want to be able to add a current pass transistor to the setup in the future, but I must test several phone type car charger circuits and their components first. I want to do all the math first to be sure the common components, most people can find, will be able to handle the design. I'm trying to avoid telling everyone how to do a massive amount of math to figure it out for themselves or give them a "paint by number" list of junk to buy. I want to distil the information down to its bare necessities first. I'm also using an analog Voltmeter upcycled from a cheap multimeter, and an ammeter upcycled from a car battery charger. I'll give references to how to reuse them. Once this supply is complete I'll be able to demonstrate how to reuse them myself, as this project is able to adjust down to zero volts or even below zero if you want. Judging by my current situation I should have this project up before next weekend so long as my back stays in decent shape ;)
I have to try this
How to make 8v to 60v power supply from 12v 20amp china battery charger...
Please make a video 😢
Make tutorial adjustable psu with powersupply pc , make 24volt dc high current please
Nice. I will try...
hello, so if i would calculate R1 and R2 to get ~12V and replace those 2 SMD resistors with 2 other SMD resistors i would set it permanently to 12v without any limitations (if the power supply was i.e.19v) ? or do i need to be carefull if i draw more current?
12 Volt supplies are the most common ones I've found in general. I would use one of them personally...
With a peak mode controller you'll be lowering the 2nd half of the wave basically making more current ripple noise. It might stress the capacitors more depending on how or if there is an LC output filter on them.
To answer your question, yes, it's just a matter of calculating resistors. However the equation is somewhat simplified. I would recommend you to replace one resistor with a similar value potentiometer. Then use the potentiometer to dial in your desired output. If you want to use 2 resistors only, measure the potentiometer's resistance carefully and match it with a resistor.
It just has to be a simple peak mode controller Circuit for this to work. If your trying to mod one like the old Dell laptop power bricks with well over 5 amps of current, you'll probably run into a PWM controller that is way more complicated. Those are about like ATX supplies. I have one of those. It has 2 transformers inside and is way too complicated for my tastes. It has 2 output polarities but it uses earth ground from the AC mains to create them and the circuit is really odd. Any way around it this is a mod you do for a "bench" power supply. This is not the kind of modification that you do to a device you plan to leave unattended. When a SMPS is engineered everything is designed to create an intersection between the oscillator timing and core saturation regions. If you go too far away from this crossing point in one direction(lower voltages), you'll get a lot of noise, if you go too far the other direction you'll get heat/smoke.
All that said, here's a commercial example of how versatile the output of a switch mode supply can be:
i1160.photobucket.com/albums/q500/jakebikereligion/20170602_182031_zps24vdy4hq.jpg
Love your video. Very clear steps. You know what you are doing ! :-)
I'm trying to modify a Dell PA-6 Family. Do you have any experience with that model? Originally the Cathode and Ref of the AS431 are connected (Zero Ohm). I disconnected them and inserted voltage divider which increased the output voltage at the Cathode to 3V but the output of the power adapter is unstable. At power on, it goes to 18V and then slowly drops to almost 0V
Any tips or ideas?
PS: a schematic of the PA6 family adapters seems nowhere to be found
thanks !
chrisV
Why dont you just put a buck/boost converter with current limit feature module and a volt/amp meter display? It would be easier i guess
Because this is essentially free. Both options are super noisy and pretty much useless for a real bench power supply. The little '0-30 volt 0-2amp' bench power supply kits based on the TL081 are a far better option to build. There is also a kit based on an ATmega8 with a little 1602 display. I have both and have been thinking about doing a video showing them in trace by trace detail along with modifications. Both kits are based on open source projects. The TL081 kit is from an old magazine article from the 1980's and the mega8 is published online.
I am currently working on a huge AVR Transistor Tester project and don't want to get distracted though, so I don't know if I will ever get around to content about the power supply kits.
The cheaper TL081's are probably the most interesting kit. If I understand the circuit correctly, the board uses a small zener voltage reference. This small voltage is then amplified to make a full scale voltage reference that goes all the way to the max output voltage. This full scale reference is then attenuated to create the desired output. That's completely different than most linear power supply circuits.
I tried to model this circuit using the tiny "EveryCircuit" spice simulator. That app is too small to fit the whole circuit and it can't handle multiple op amp supply voltages as the circuit is designed. I've been meaning to try it in LT Spice but haven't yet.
...anyways... all I was showing with this upload is how the TL431 voltage reference is used to regulate the output just like a linear regulator's internal voltage reference works. There are problems with this circuit and it should never be left unattended. The biggest issue is how the auxiliary power supply for the switching IC changes when the secondary output is adjusted. The second biggest issue is that the frequency of the switching controller and the inductance of the transformer are specifically selected and designed for. Changing the output voltage also alters the crossover point of the design and will cause problems. The third biggest issue is that these supplies usually have an output filter with 2 capacitors and an inductor in addition to a choke. Altering the output characteristics will cause issues with this filter rendering it useless at best or even causing it to amplify the problems further. There is a margin of error built into these designs but the further the output gets from the original design the worse the problems will be. This applies to all switching power supplies including the Chinese modules.
When it comes to a bench power supply the most important traits are stability, over/under voltages, and what happens when it is powered on or off. Most switching supplies suck at these specs.
-Jake
Yeah, but sometimes opening a power circuit and plugging it in is not a good idea for begginers (and the input caps are also dangerous). Those who dont understand properly could got hurt or killed. And not all laptop charger has the same circuit, if someone actualy tried to modify their own charger, they might blew something up. Anyway, thank you for your explanation, keep up the good work sir! :)
-sorry for my terrible english
The higher the resistance on the r1 the lower the output voltage? Or it will go higher?
I don't recall, and am not rewatching, but the first adjustment pot sets the high voltage limit. It's the one I gestured to show: "once it is set, don't touch it again." The second pot adjusts the output.
The lowest voltage setting is something you can't control unless you (re)design the circuit. It has to do with the auxiliary coil on the mains voltage side. The aux winding ratio/rectification/cap are usually the main issue. When you adjust the output of the secondary, you're also adjusting the output of the auxiliary winding. Once this is low enough, the switching circuit loses power and the control chip shuts down. You need to completely understand the feedback circuit if you wish to modify the aux circuit as these two circuit blocks are integrated on many supplies. Watch a bunch of videos from DiodesGoneWild where Danyk explains switching power supply topologies in depth. That's the main place where I learned.
There is a lot of potential for issues, and "beware of deadly mains voltages" should be quite obvious here, but that said, it is possible to use a second isolated power supply in place of the aux winding on most power bricks with a controller chip. In theory, this would allow you to adjust the supply down to whatever the control chip's internal reference voltage is.
Also, I shouldn't need to tell people this, but many seem to need it: modifying the circuit as shown is intended for simple bench use. Never leave it unattended, or attempt to use this with high loads for extended periods of time. The original as-labeled output should also be considered "ideal." Derate the output by the percentage change. If you adjust to 1/2 the original output voltage, assume you've cut the efficiency by 1/2, doubled the noise, and halved the maximum output wattage.
No matter what you do, the wattage rating can only go down. This limit is baked into every component used in the circuit design and can't be changed. For instance, a 10 watt supply is never going to be more than a 10 watt supply. If it is set at 10V, 10 watts is 1 amp. If the output is 20V, 10 watts is 500mA. You'll never be able to violate Ohm's Law.
-Jake
@@UpcycleElectronics do i need to change the orginal value resistor ? Gnd to ref pin is 10k and ref pin to + is 60k+500ohms. All i need is to change the 60k resistor with 60k or 50k pot right?
@@smoke_stackz3168
If you do that, you won't be able to control the voltage as shown. The idea is to set your (main/second) adjustment pot to maximum resistance, then set the resistance of trimmer pot (the one to ground), as shown, until the maximum voltage (with an extra margin) for your capacitors is reached. If you skip this step, you're probably going to blow up your capacitors, and you should check the voltage ratings for everything on the secondary side too. You're probably going to burn/blow up everything as you haven't set the output maximum, but you might get lucky.
@@UpcycleElectronics ur right i replace the resistor 10k to 4.7k and replace the 60k with 20kpot+4.7k and the output is pulsing at 13.5 to 13.8v looks like the adapter is turning on and off and led is blinking slowly
@@smoke_stackz3168
Some times they need a load on the output to work at a little bit lower voltage. Usually this means the aux winding isn't making enough power though.
Nice job, i got 12v 120w Dell laptop Power adapter, can i make from it 19.5v 240w adapter?
Thanks for the comment. Sorry but...no you can't make it more powerful. You need to learn how inductors work. You can't make the transformer's saturation region change by changing a resistor. The power will go down the further you adjust the output away from the original design. The power factor will also change, as will the efficiency. This mod is intended as a simple way to create an adjustable bench power supply. I don't recommend trying this for any power supply that you leave unattended. There is an open loop protection circuit on the primary side of all UL listed "off line" (AC mains) switching power supplies that should prevent the output from getting high enough to cause it to catch fire because of oversaturating the transformer. If you want to create a higher output it requires a lot of math and an understanding of high frequency transformer design. The supply you have will be designed with a specific turns ratio on the primary and secondary windings. This will also match the cross sectional area of the core. All of these calculations relate to the switching frequency set by a special capacitor and several other factors depending on the design of the switching IC. It gets really complicated to understand and work out all of the math. However after everything is calculated for a circuit like this, the designer of the circuit is not creating a precision device. The final circuit design has a considerable amount of wiggle room. I've simply shown how to take advantage of this "wiggle room" to create something useful.
-Jake
I got another one with 19.5v but 120w instead 240 and amperage is little lower. Will i have better chance in this case? Thanks.
FREAD
Go to the app store and get an Ohm's law calculator app until you get a solid grasp of how Ohm's law works.
Power(watts) = Voltage × Current (amperes)
It's usually just written P=VI
Watts, Current (Amps), Voltage, and Resistance are all related. If you have a circuit with a certain Voltage, and Current already, then you decide you want more voltage, no matter what you do to the voltage it is not going to add more energy into the circuit than what you started with. So if you can't add more energy, how do you change the voltage? Simple, you must reduce the current. This will always happen whether you expect it or not.
If you have a power supply that is rated for 120 watts, that's the total energy available as that power supply is designed. You would need to totally redesign the entire circuit from scratch to make this number higher.
That said, if your looking for a 240watt 19 volt supply, your looking for around 16 amps of current. That's getting into industrial grade hardware. A regular household circuit breaker can't even supply that much current. You'll need a design specific for that kind of output. Once a DC converter type power supply gets past around 100 watts the circuitry needs to incorporate power factor correction. PFC is complicated and I don't completely understand it myself. It's a circuit that prevents the supply current from lagging behind the peak voltage of the AC waveform. It has to do with efficiency and how much noise the circuit adds onto the mains AC power line.
What I showed here will only work on simple peak mode type circuits anyways. If you open up that Dell you will probably find a completely different type of circuit. The type of circuit I showed here are the most common ones for laptops and printers and they usually are rated for 12 to 36 volts and 2 to 6 amps max. Once the current goes higher the design usually changes to a PWM type of controller. They are more flexible for high current. The main difference is that they have 2 voltage feedback circuits. There's one loop for the voltage and another for the current. You can't modify those with potentiometers because both feedback loops are connected in how they cross each other. If you just change one feedback loop, for example, the voltage feedback only, the circuit will fail quickly. This type of design is used in almost all PC power supplies like the ATX. When people try to make bench supplies with them they tend to fail because of how this feedback system works. They are much harder to modify and understand properly. The most common IC used for that type of supply is the TL494. There are many circuit references online for how to build a PSU around that IC. There are probably a few calculators available too if you search for "TL494 design calculator." That's the IC you will probably need to use or figure out if you want to build something as serious as your 240W@19V/16A setup.
-Jake
Thanks a lot, probably is better buy new but laptop is old Dell Alienware and doesn't worth to spend on it anymore.
use 3055 power transistor at output for better results
Yeah. I've actually played around with this circuit again recently on my bench. I've been trying to figure out different combinations to create a better adjustment range. Just using a power NPN creates too much heat for the enclosure and heat sink.
As an aside, there is usually a spot for a second TO-220 rectifier in most of the generic laptop PSU's. I have been playing with ways to make use of this by cutting a few traces. This allows use of the heat sink already installed, and the potential to keep everything in the original enclosure.
The difficulty for me has been finding a way to track 2 separate adjustments, one for the SMPS and the other for the npn pass transistor. I've been playing with ideas. I think I need to use an op amp to effectively control both circuits with the appropriate offset. I think I need to order a small dual potentiometer though. I want to try a circuit using 2 separate resistor dividers and a zener to set a minimum threshold for the SMPS.
However I have also done a good bit of experimentation with changing secondary windings and ratios on flybacks recently. The best SMPS hack may be adding a small additional secondary winding. Just a few turns is usually all anyone needs for a few volts. This winding can then be shunt regulated for stability and used to power an independent adjustment circuit using a couple of scavenged 431's, a power NPN and a N-FET. This can create an adjustable voltage and current. Unfortunately this probably won't fit in a factory laptop power brick unless you get super creative ;)
-Jake
had a 4amp 16v supply i hacked it to 19v supply but iam getting lower current output and voltage fluctuating when load is connected
what should i do to rectify this problem
seems like its working frequency is reduced
please help
Your drawing too much current then. Your probably saturating the transformer or tripping some kind of protection circuitry related to duty cycle. The ideal calculation of ohms law for the information you gave is simple, 16V@4A= 64W. If you reverse this you can say your power supply is capable of 64w when operated as it is designed. Next you can calculate the current of 19 volts at 64 watts just to get an idea of what the perfect world number is. That comes out to 3.3684 amps from 19 volts@64watts. Of course this will never actually happen because a SMPS is designed for a specific crossover point between the oscillator timing and the flux density characteristics of the transformer. The farther we deviate from the crossover point the worse things will get. Personally, I would assume I could achieve no more than 70%-80% of the ideal number.
I specifically described how to build a bench power supply for simple bench testing in this demonstration. Never leave something like this unattended, especially if your using high current at a higher voltage. If the device is compliant with regulations in most countries the design should be limited by the optocoupler and circuitry on the primary side before it gets into a catastrophically dangerous condition. If the optocoupler or any part of the secondary feedback loop gets damaged, the output of the controller will peg out at maximum. The voltage reference inside the actual SMPS controller chip will be compared to a zero instead of the feedback voltage it normally "sees" from the primary side of the optocoupler. Any safe design should account for this condition. The largest physical limitation is the duty cycle of the controller and saturation of the transformer.
If you want to push SMPS to its limits you really need to fully understand how it works. It's not simple. The transformer equations are probably the biggest challenge. That's actually where I'm at with my own learning. I'm not an Electrical Engineer, I'm a hobbyist like you. I read a lot and try to share what I've learned. I started a series on building one of the transistor tester clone tools because I need to measure inductors before I can take my understanding of SMPS circuits to the next level.
I have been saving a lot of good references on the subject in my 'Power Supply' playlist on my channel. The hardcore SMPS stuff is towards the bottom. There's a few practical examples with transformer design/winding from Silvio de Leonardo, and the theory stuff from Sam Ben-Yaakov is mind blowing. He has a video posted on everything SMPS engineering related. You also need a thorough understanding of MOSFETs and especially their gate capacitance effects. I think the best possible reference is from the Free MIT Open Courseware channel, specifically the 6.002 classes and lecture 9 a and b of that course. I hope you understand I am not trying to discourage you, or talking down. I'm no expert. I just rea a lot and watch a bunch of technical stuff. These references are not something I'm telling you to do. They are literally what I am currently working on understanding myself.
The latest SMPS project I got 3/4 of the way through before getting a component tester has seen me smoke 3 IC's, a trimmer pot, and a FET all in the name of exploration and learning.
Be careful, and best of luck, -Jake.
Upcycle Electronics
i am goint to learn about the controller ic and its working
so that i could figure out what is happening
thanks
Silent but wonderful.
Great video, thank you?
Man, I have a power supply here, and instead of a TL431, it have a OP-AMP LM101A. Did you ever found one with this kind of feedback circuit? I'm trying to understand it to hack...
The TL431 is essentially the same. Inside, it is basically just an op amp configured as a comparator with a zener reference. The modification I showed here is how to change the feedback loop of the op amp. Most op amps have trouble handling the same amount of output power as the TL431 but on the equivalent circuit in the TI datasheet for the TL431 it shows an NPN transistor on the output of the (equivalent) op amp. A similar transistor configuration on any other op amp should work the same. That isn't really needed though in this application. Most op amps can power an LED sufficiently. That is all that is happening in this circuit. When the comparator reaches it's threshold voltage it switches on the LED inside the optocoupler. Then that signal is transmitted through the plastic to the phototransistor/reciever on the other half/primary side. Technically this comparator circuit could be done discretely but creating sharp signal transition edges would be more challenging.
I looked at the summary specs on TI for the 101a. It looks like that op amp is designed for temperature stability. It also has a larger input voltage range than the 431's. It is unusual though, because the 431's are practically cost free in large quantities for production. The 101a would have cost several cents instead of a fraction of a cent for a 431 not to mention all the space and configuration components. The 431 is even available in a SOP8 package like an op amp if the package space were an issue. Although the most common 431 is probably the SOT23 version which wins the board footprint size/space battle every time.
My first thought was that they may have had some kind of dual op amp feedback loop, but after looking up the 101a, it is just a single channel op amp. It's definitely odd, and something I haven't seen before.
-Jake
A really simple fundamental explanation of the circuit is that the optocoupler is just a switch. As soon as the controller 'sees' the switch has been flipped on it holds the output voltage at that point.
If you think of the moment the switch is flipped like the line on a road, the controller want's to follow that line as closely as possible. It will have a tendency to go barely under that line but is much less likely to go over it. When something changes, like when a load is connected, in a fraction of a second the voltage will drop due to the change in current/resistance. This will turn off the optocoupler switch, and the controller will adjust the duty cycle of the signal on the primary side to compensate until the switch is flipped on again. All that is happening on the secondary side is a circuit that tells the controller "👍 or 👎"
There are multiple ways of designing this, but in the end this is a design that must create a yes or no feedback...so long as there is only one feedback channel/optocoupler. If there are two like a TL494 or KA7500 it's a totally different thing.
@@UpcycleElectronics the operation of the optoacopler is like on and off, and not by the intensity of the led so?
@@UpcycleElectronics in this case, the opamp is operating as a comparator.
@@sigmawarrior.fokeryou
Yeah it's just an on off switch.
Another general rule. Op Amps are designed to create an extremely large amplification factor or "gain" from input to output. This is needed so that they are very predictable with a feedback loop. When a single transistor is used in a circuit as an amplifier it has much lower gain. If you want to build something analogue based on a single transistor like what is inside an optocoupler, you'll need to tune and bias the circuit for every single device made because no two transistors will operate exactly the same. If you look at some devices from 30-40+ years ago you might find a few circuits like this. They will have lots of trim pots, matched pairs of transistors, and other tricks to make complex circuits that just were not available as integrated circuits back then. Nothing is made like this any more.
Any design you will find now employes circuits with near infinite amplification and a feedback loop to control the amplification. These types of designs do not rely on properties that are subject to variations between individual components.
So what does this mean? Any time your looking at a circuit like this where there is a single transistor and/or LED, it is safe to assume they are operating digitally as an on off switch. There is no way that these could be used reliably to transmit analogue information in a mass produced product.
-Jake
Sir this tutorial is extremely informative. But due to my lack of knowledge of SMPS principle. I wanted to ask you, i have a 14 volt 3.6amp psu. I would somehow change the voltage to 5 volts by the way you explained. But is there a way to increase the current output.
I waant to convert this 14 volt 3.6 amps psu into 5volt 6amps atleast.
I am planning to make a multiport usb charger out of an old psu lying around. If this cannot be done with this 14 volt psu, please suggest me some other ideas or guide me how to achive 5 volts with atleast 6amps
Please guide me sir.
That's not possible without designing a supply from the ground up.
@@UpcycleElectronics so the only option is to lower the voltage by the way you explained?
@@anirudhkundu3567
There are many many issues. The biggest one is that the original watt rating is the best case scenario for the power output as designed. As soon as you change the output voltage, you also take away the efficiency, and therefore the wattage rating will decrease. SMPS design is all about taking several different variables that are about like dots spread out around a circle and making them cross over at a certain point. There is some margin of error and flexibility available at the crossing point. The points are the switching frequency, controller circuit design, and the transformer properties/windings ratio. Those are just the main issues there are more too. If you change the voltage it will impact several variables. Nothing is free here. If you change one thing a little bit it will affect the other things a little more. You don't need to understand the entire system. It's outputs will operate accordingly. If you want to change the voltage 120% you should expect the current to go the other way by about 150%. By other way, I mean in the negative direction. The way you don't want it to change.
I demonstrated how to make a simple "bench power supply." Not how to hack and modify a SMPS for use as that's not even possible. This is intended as a tool that is never left unattended or stressed near it's limits. This is a cheap way to make something similar to a LM317 bench power supply. If you want to build a SMPS for some specific application, you need to take the time to learn and develop a comprehensive understanding of power supply design. I can't help you with that as we would be like two blind people walking around on a mountain. I'm no expert at this stuff. I just shared something simple that I learned here ;)
Good luck.
-Jake
@@UpcycleElectronics can it not be done by modifying the switch mode transformer by rewinding the transformer and adjusting the feedback resistors?
@@anirudhkundu3567
You would need to calculate the entire circuit to accomplish your goal.
I made it !!! but it adjustable to 5v3 to 12v, from 5v DVD PSU, any suggestion for lower setting (to 0v) ?? *need help
You can't safely go lower with this kind of design. You need to add another circuit. This is why I uploaded the second power supply video showing how to mod a phone charger circuit.
The only way to make a mains power brick go lower is to mod the primary side and that absolutely requires a complete understanding of all of the engineering and math involved. If you were to mod the primary side to go lower and then tried to adjust it for maximum output there is no doubt something WILL get smoked or worse. There are a few ways to make your own SMPS devices that have more adjustment range but they all require more circuitry. I'm building the transistor tester because I need to measure small inductors before I can detail more complex SMPS designs. I have one SMPS project 3/4 complete at over 15 min of fully edited footage and another 2 SMPS projects where I've built a prototype to enable me to design a video around.... More to come ;)
alright .. thx for all :)
I've built second one, it is adjustable 5v1 - 19v from 19v Laptop charger, but no easily to adjust @16-19v & smooth adjustable on 5v1 - 12v, any suggestion?
TheJavaSync
The only thing I can think of is your potentiometer has a logarithmic taper scale instead of linear taper.
alright, it seems diff potensio, I'll looking for linier one, then. Thx U very muchhhhhh !!!
Thanks
I just don't understand what did you do with your dmm, what you measure and what it shows it seems doesn't work?
Can you add a constant current feautre to this hack
Foldback current limiting would be easy to add. Just use a Power NPN transistor on the ground side.
Instead of connecting your load from the positive voltage to the ground terminal, you add the Power NPN and a 500ohm potentiometer between them. The ground side of your load connects to the Collector of the NPN. The potentiometer is connected like a rheostat (leave 1 end connection disconnected). Connect the other side of the potentiometer to the Collector of the NPN. Then connect the adjustment wiper terminal to the Base of the NPN transistor. Lastly connect the Emitter of the transistor to the old ground terminal of the power supply.
This is the simplest Current Pass circuit. The main drawback is that you lose a diode drop voltage across the transistor. This doesn't matter if you can adjust the voltage of the power supply itself. The diode drop is typically around 0.7 volts. If you use this circuit with a fixed supply like a 5 volt supply, you would only have around 4.3 volts to power the circuit. Like I said, if you can adjust the supply this doesn't matter.
The other way to think about this NPN + Potentiometer circuit is to think of it as a Power Potentiometer with a Current Bypass transistor. You could always buy a 100W Power Potentiometer with a resistance in the range you're looking for. These cost a lot. The circuit I described will even work with a trimmer pot (although these are not intended for regular use and adjustment). You can build this with almost any Power NPN the classic 2N3055 works fine. Just make sure you use one with a high enough current and voltage rating.
Build a couple of these, they are great adjustable dummy loads for testing lots of stuff.
hope that helps :-)
-Jake
Upcycle Electronics thanks jake and i know the method with the npn transistor and that’s just waste of power it will be better with an op amp but i was asking if you could make varible constant current inside without any parts outside
Mary labtop Ka charge krent rok rok kr dyta ha battery b charge ni krta koy jogar bato lead b change kr d ha
Thank You, you just got new sub
Nossa que ideia genial!!!!!!
Obrigado por assistir! ;-)
what is the minimum and maximum safe voltage with this method ?
It depends on the circuit. The part of the circuit that determines this this the auxiliary power winding on the primary side.
When this type of power supply is plugged in, there is a high value resistor that provides just enough current to get the control chip started. Once the chip starts working it draws full power from an auxiliary winding, plus rectifier/diode and a small filter cap. Sometimes this part of the circuit will also have a zener diode to provide additional over voltage protection.
The auxiliary winding must be designed so that it produces the right output power when/while the secondary output is running at the desired power. There is usually a lot of room to mess around with this part of the design. These chips usually have a supply voltage range from around 4 to 30 volts. Most of the time the auxiliary winding is designed for the absolute maximum efficiency point of operation which is somewhere around 10-15 volts. This is done because the power lost in this part of the circuit is the primary cause of idle losses when no load is connected. When you lower the secondary output, eventually there will not be enough power coming from this auxiliary winding and that will create the low voltage limit. This circuit is sometimes replaced with an additional power supply.
If you understand mains supplies well enough to do a thing like this, the low output limit will just be the control chip's internal reference voltage.
The upper limit should be set like I showed in this upload. Don't attempt to exceed the power rating marked on the power supply already. Power (watts) = voltage x current. If you raise the voltage you must lower the current, but the power must remain the same. You should also assume the power output that was originally designed into the circuit is going to be the most efficient point of operation, the further you get from the original output, the further you should de-rate the power. If you want a 20% higher output voltage, assume the totally power will need to decrease by around 20% due to inefficiency. This should keep you from saturating the magnetics, at least in a simple design like I've shown. Still I wouldn't recommend leaving such a thing unattended if you make large changes to the output.
I still have and use the one I showed in the beginning. It works well from around 7.5 to 30 volts. I've never left it unattended, but it's handy for quick testing of circuits. Hopefully this helps :-)
-Jake
@@UpcycleElectronics Thank You :)
What multimeter did you used in this video?
Strefa PC
I wouldn't recommend it for electronics. It's an "Ideal" branded meter I got from a big box home improvement store here in the US many years ago. The resolution is terrible. You can get a far better meter for the same $60-$100. This is the kind of meter you use when wiring house Mains or in an industrial setting where you want to measure 240v or 480v.
If your looking for a meter for electronics the best value is the Aneng 8009 aka the 9999count meter. There is also another version branded as a bSide Z302 (9999 count) meter for the same price but is produced in smaller quantities and takes a bit more searching to find. If you shop around (AliExpress) you shouldn't pay more than $25 to $30 shipped for this meter. For around $40 there is a new Aneng meter that has a 20k count. It was just released a few months ago and is another remarkable value.
If you are bent on a meter you can be confident using with high voltages AND use for electronics, go check out the ones Dave Jones designs and sells through his EEVBlog website.
If you want an outstanding bench DMM (not a handheld portable) get on eBay and find a hideously ugly Keithley 197 for less than $100. I paid just under $100 for mine ($80 + shipping). That's a well documented 200,000 count meter.
I think I paid around $80 for the Ideal meter ages ago. IIRC it's a 1999 count meter. It's definitely not a good meter for electronics work. I got it long before I got interested in the hobby.
-Jake
Sound vid + description all the way around..
Cómo puedo contactar con usted
Where do you find these rotary potentiometers?
Old 5.25" DVD drives.
Here I am thinking the days of silent movies were over with.
Sir can you make HP 0957-2242 AC Power Adapter 32V adjustable ?
You tell me. I've shown how I do this. Just follow the steps as shown.
@@UpcycleElectronics do you have any page or something that i can send pictures of it?
@@zedthekiller3083
no.
@@UpcycleElectronics okay sir i have problem about number 5 about the TL431
@@zedthekiller3083 ?
excelent .
Skema Sir. . .
10 to 25 vdv for smps uc3843
My list of power supply references I have saved:
th-cam.com/play/PLPIwHuVy9EyOtJ_LYw_Rc8oFvBd5TUTYN.html
Upload explaining a UC384× circuit:
th-cam.com/video/B19rB_FR5Mk/w-d-xo.html
Danyk's hand drawn schematic from that video:
danyk.cz/reverz44_en.html
-Jake
How to adjust to 0volt - 20 volt? I need calibration 0-19 volt for power supply handphone
i misknow,, please attach circuit smps. thanks
What is your language?
my language is a persian.
این یک آموزش DIY است که نشان می دهد چگونه من منابع برق ارزان را تغییر دهم. من از 5 مرحله ساده برای پیدا کردن و جایگزینی 2 مقاومت با پتانسیومتر استفاده می کنم. . این برای من کار کرده است من از لپ تاپ و منبع تغذیه چاپگر استفاده می کنم.
این کار را انجام ندهید، مگر اینکه خطراتی را که در هنگام استفاده از خازن های ولتاژ بالا و ولتاژهای AC درک می کنید را درک کنید. شما مسئول ایمنی خود هستید. یک اشتباه کوچک می تواند شما را بکشد. اگر خطرات را کاملا درک نکنید، آنها را نگیرید. هنگام تنظیم ولتاژ نهایی حداکثر باید از امتیاز ولتاژ خازن خروجی اطمینان داشته باشید.
اصلاحیه ها:
من 2 مقاومت از منبع خروجی DC منبع تغذیه را تعویض می کنم. این دو مقاومت یک تقسیم ولتاژ ساده ایجاد می کنند. تقسیم ولتاژ به یک تنظیم کننده شونده zener متصل شده و خروجی آن به یک optocoupler وصل می شود. شما لازم نیست که مدار را بدانید و یا آن را درک کنید. optocoupler همیشه یک تراشه خم شدن 4 پین است که عبور از مانع انزوا را در زیر ترانسفورماتور. معمولا برای تقسیم ولتاژ و برخی دیگر برای مقاومت در برابر ولتاژ و یا پین ولتاژ ورودی رگولاتور شنت، معمولا دو مقاومت وجود دارد. نگران نباشید. خود تنظیم کننده شنت همیشه TL431 است
تنظیم کننده شنت یک اتصال دارد که خروجی کامل ولتاژ دارد (یا نزدیک به آن (بی طرفی))، اتصال دوم در زمین است، و آخرین اتصال در 2.4-2.5 ولت (Vref) خواهد بود. فقط به کلاههای خروجی نگاه کنید تا نشانهای مثبت و زمین پیدا کنید.
معادله ساده برای خروجی بازخورد است
Vout = (1 + R1 / R2) * Vref
من نمونه ای با مقادیر این ویدیو نشان داده ام. این معادله به 12.44 ولت رسید. اگر به دقت تماشا کنید، در واقع 13.44 ولت دریافت کردم. این به این دلیل بود که من نمی خواستم همه چیز را با پیچیدگی نشان دهم که چگونه یک تفکیک جزئی کوچک در پین ولتاژ ورودی TL431 وجود دارد. اگر متوجه شدید 12.12 ولت قدرت خروجی به دلیل مقاومت بیشتر از آنچه که من با پتانسیومتر جایگزین کرد، نبود. تنها دلیل ذکر معادله این است که مطمئن شوید که خازن خروجی (ها) را به طور تصادفی منفجر نکنید. دقت معادله مهم نیست، فقط تقریبی است.
مقاومت می تواند شناسایی شود، R1 بین ولتاژ خروجی و پین Vref متصل می شود، R2 بین پین Vref و زمین متصل می شود. 2 پتانسیومتر را انتخاب کنید که نزدیک به مقادیر این مقاومت است. مقاومت الكتریك R2 باید یک گلدان حرارتی باشد زیرا شما فقط برای تنظیم حداكثر ولتاژ خروجی به آن نیاز دارید. پس از حداکثر خروجی، از پتانسیومتر شما با R1 جایگزین کنید تا ولتاژ خروجی را تنظیم کنید.
محدودیتی برای ولتاژ خروجی بالا و پایین شما با هر طراحی SMPS وجود دارد. این کار با مشخصات ترانسفورماتور، تراشه کنترل کننده فرکانس و زمان خازن انجام می شود.
اکثر منابع تغذیه ای که من تغییر دادم از حدود 7.5 تا 22 لیتر ولتاژ پایدار است. عرضه می تواند بالاتر یا پایین تر از حد متعادل باشد، اما هنگامی که یک بار استفاده می شود، آن را می کشد.
تنظیم زمان اول:
ترفند برای ساختن تنظیم برای اولین بار این است؛ ابتدا، معادله را انجام دهید تا ولتاژ حداکثر را با هر دو پتانسیومتر تعیین شده در مقاومت کامل بدانید. دوم، بررسی کنید که هر دو پتانسیومتر به عنوان رئواستات تنظیم شده اند (تنها 2 از 3 پین به سیم کشی هدایت می شوند)، و آنها را به ارزش مقاومت کامل خود تبدیل می کنند، یک متر و این را بررسی کنید؛ سوم، روشن کردن، و به آرامی تثبیت کننده ترموم (R2 / 5K پودر ترمور در این ویدئو) را در حالی که تماشای یک مولتی متر متصل به خروجی را کاهش می دهد. حداکثر ولتاژ خروجی خود را تنظیم کنید. پس از این تنظیم، هرگز این پتانسیومتر سهبعدی را لمس نکنید. تنوع خروجی شما کاملا توسط پتانسیومتر R1 از این نقطه به جلو کنترل می شود.
Does that help? This video is an example of how I do a modification like this. If my example doesn't make sense visually, you probably shouldn't try anything like this. I've simply tried to show how these power supplies operate just like any other adjustable regulator. There are big limitations and compromises made when adjusting the voltage on a device like this. My intention with this video was to show a different way of thinking about voltage regulators and how they work. I do not want to create a schematic or the detailed information. The person that does not understand the fundamental information I demonstrated here should avoid a project like this. These devices are very dangerous.
woww, persian type ,,
ok, thank you.
muy bueno tienes tu like
how many amps does this put out?
User 820
There is a simple and complicated answer.
The simplest answer is this example:
lh3.googleusercontent.com/-AJRuQwP_BQg/WZB87RRtAxI/AAAAAAAAAaQ/KsJZt0lEyCkKjZMXmY4RoTXv9-teNUKQQCHMYCw/s1600/Screenshot_2017-08-13-09-20-04-1.png
I am comfortable using the current rating of my supplies at the original voltage rating. I have used about 70-80% of the rated max current of a supply within 5-8 volts of the original voltage of the supply. Once the output voltage gets farther from the original design you have to delve deeper into the design to find it's limits. The timing capacitor and transformer design need dramatic changes in order to create several amps of current reliably. My supplies that I have modified have no problem supplying 1-2 amps of current at higher and lower values, but I have not pushed them to their continuous voltage/current limits. I have tested brief outputs with low value resistors and seen 3.4A at 22V from the supply I showed modded in it's original enclosure. The frequency and duty cycle of the output transistors could become a major issue if you push them continuously outside of their original design parameters.
The values used for the resistor divider have no direct effect on the output current. They can be anything from 1k to about 100k. It's the ratio relationship that is important. I used 5k and 20k because they were close to the original resistor values and it helped simplify this example. They could have been 2k/5k, 2.5k/10k, 20k/50k, or any other combinations that can achieve a similar 1:4 ratio.
All of the controllers I have shown, and indeed most cheap SMPS controllers used in laptops/printers are called "peak current controllers." This type of controller varies the frequency of the AC signal to change the output voltage.
I can give you a lot more info if you really want to know the grity details, just ask.
Fortunately, the method I showed here only biased the secondary side of an isolated SMPS. There is another bias point on the other side of the optocoupler (the 4 pin IC). This limits the potential for someone to really do damage or blow up a circuit.
If you have a specific project you want to power at higher currents it's best to start with a SMPS at close to your desired output. For general bench projects my supply has yet to reach its limits.
I will be showing another MC34063 example soon (current project I'm working on) that is part of an AC to DC design anyone can build. I'm trying to break down the details even more and show some of the limitations and functionality of a simple adjustable SMPS. I will try and demonstrate the characteristics on my oscilloscope once I finish the project enclosure build. I plan to demonstrate the design from this video, the current project, and compare it to a cheap commercial LM317 supply I have. That should help clarify the subject a bit....
...and this is why I don't speak in my videos to date.... Describing electronics is a nightmare....
All of that said, I've read a ton, and experimented a lot, but I'm not an electrical engineer. I make mistakes. I am not an expert. All I can tell you is this is what I've experienced and seen for myself. I have seen repeatable results and shared them as such, but I can't represent them as anything more :)
Upcycle Electronics thanks! :)
Around 2 - 2.5A max.
sees 420v *dabs*
What is the minimum output voltage?
genuino 301998
It depends on the controller and how it is configured. Usually you can't get below 7-8 volts on a device originally designed for 12-18 volts. There are several factors that cause this. It's mostly safety related factors built into the design to prevent catastrophic failures.
how many power? (Watts)
YouTubiando
The maximum power output should be whatever the unit was originally rated for at the rated output voltage/current. As you get further from this number the efficiency will decrease and you will get less total output.
First of all, congrats on the unique idea and the fantastic job and thanks for taking the pain to share it. My apologies if you are vocally challenged, but if not, by going mute you have unnecessarily complicated a subject that is anyways not easy for a non-electronics person to understand. I completely lost you in the part of the video that has the number 1-5 annotations. The great pain you took to write it all down in the description section does not help at all because it is not the same as a verbal commentary or text annotations explaining the steps as these are being performed.
Howdy
Upcycle Electronics c
percuma adapter seperti itu dibuat variable, kalau cuma sekedar merubah besaran output bukan hal yang sulit saat diberikan beban protek aktif sehingga adapter itu tidak bisa jadi serba guna so for what. cobalah buat konten yang merubah adapter menjadi high load atau menghilangkan auto protection supaya bisa sesuai judul high current.
your multimeter was showing 7.5 volts before you turned your power supply on... you got bad diodes.
Walter Bunn
7.5mV and unless I'm mistaken (quite possible) that's just noise on the leads open circuit.
Ah... no i realized what it was...
i figured there was an "on" switch with the potentiometer, and there is not.
It's just on if it's plugged in, which the multimeter correctly recorded.
It goes from 7.5 volts to 27 volts unloaded... s'alright.
استعراض العضلات
Needs NARRATION otherwise it's a MIME show.
can you give me your tester ? :)
mucho enredo.
bien
Sprich doch bitte mit uns!!!
like your idea but I'm not that much expert in electronics so can you send it to me
Si hablará y explicará sería bueno ... pero no somos sordomudos ... y estoy más que seguro que si un sordomudo hace este video ..se le enfocaria el cerebro a poner un traductor con señas ...
दिखाई नहीं आधा अधूरा है वह पराई लगती हैजब कोई संबंध मीटर रनिंग कम्पटीशन दीपावली
Sigh. these "I insist on no talking to the point where I slowly make note with pens IN A VIDEO?" are so tiring as a format. Seems like if you aren't going to say anything extra just provide a set of notes or a generalised schematic. You seem nice, knowledgeable and everything, but this format is annoying.
Why is it now nobody likes to speak on there own video. Before the only people who didn’t speak were the one who didn’t want to get in trouble. But these guys are on the up n up I don’t like when they don’t explain what there doing maybe it’s because I’m slow who knows in any case i don’t like it but good videos but can be so much better
tienes las manoa de niña ja