What a fantastic day and age that we live in where generous individuals like you can teach through tutorials like this. Thank you so much for taking the time to do this, especially the theory explanation, I absolutely love it. Keep doing what you are doing!
EEVblog I am an almost complete newbie and looking through LDOs to build my own power supply (like real men do). I saw your new rigol PS and I figured I can code the software for one just like it, with color display, a nifty graycode encoder and all. I came across the same part and pretty much came up with the same schematic, except for the buffers which are an awesome fix I could not figure out (thank you for the opamp explanations as well. I finally understand the t-shirt). This tutorial is by far the best I've seen, for people like me, as it explains the train of thought that went into designing this which is far more important then etching techniques or explaining how to drill holes in aluminium. I also wanted to get more then 1 amp from it so what I did, since the part is based on a comparator, I put a PNP from vcc to output and it is opened by the _IN_ pin. Now I can get it up to 10 amps with proper cooling and a huge transistor. Spice shows some 20ma through the IC for a 3 amp output so I suppose I can get it up to 15-20 amps no problem, with an even bigger PNP and more cooling and it's still dirt cheap.
Well, generous... He was getting paid pretty well already by TH-cam when he shared this information. Don't get me wrong, I have nothing against David or this channel, I'm learning a lot from his videos too. But you have to be aware there's not just generosity at play here.
@@ArumesYT youtube doesnt pay very well, especially at the sort of range of veiws that he got at the time of your comment. his time could be more lucratively spent elswhere, so although he does get compensated for this, his decision to share this knowledge is definitely partially out of generosity.
@@ProteinFromTheSea So I say there's not JUST generosity at play. You counter that by saying it's at least PARTIALLY out of generosity. Why counter my comment when you agree? And Dave already was a full-time vlogger a this point, so apparently the pay was enough.
I have 40 years of electrical engineering experience . Just out of curiosity I looked at your video. I must admit it is an excellent presentation. I liked very much the ingenious way you build the presentation with all the "twists" in the design. Thank you very much for the brilliant presentation.
i really can't express how helpful and useful this tutorial series is! Thanks a million for describing in so much details, it's like a 2nd year uni stuff for someone who never did Electronics beyond college. Love it!
@scientist375 Sure I could drop a few minutes here and there if I used a script or went and re-shoot some parts to make them tighter, but it might get 40min down to say 30min at best. You can't magically compress 40 minutes worth of material into 5 or 10 minutes and keep the same content and explanation. I could make this video 1 minute long and say here's the schematic I prepared earlier, but what's the fun in that. The next video will be 20-30 minutes of looking at datasheets and tweaking
I personally much prefer proper, long form videos like you tend to make. It allows you to explore as in depth about the subject matter as you are so inclined. I love that, tbh.
Dave, I tip my hat to you. I've been pulling my hair out trying to work out how this is done and why the adjustable voltage is so simple yet the current limiting looks so complicated, but you have really spelled it out in a very descriptive way. Thanks so much!
@zox012 Yeah, I know. It was supposed to be one episode and I shot most of it thinking that, but when I got to editing, as is often the case, I was surprised at the amount of footage I shot. It was far too long, and kinda made sense to split up into a couple of episodes. So it should be at least 3 parts now.
Looking forward to the rest of the tutorial. I like how you evolved the design from using more components to a little more elegant with fewer components.
@scientist375 And the Makezine videos are quite lacking in detailed explained content. If you like that, fine, but it's not the way I like to explain stuff. I chose to show you some of my thought processes in the design, and that takes time. You can't magically shorten it, there is only so much you can fit in 5-10 minutes.
@mitpatterson It's orange :-> That's just the colour of various length pre-cut jumper links I have. I am using red and green for the power rails if you look on the left side.
Your accent/voice is actually the main reason I'm here, and maybe enthusiasm about electronics design. Really great blog, honestly. May I request that you do a review about low-cost (0-80$) dc adjustable (0~30V) laboratory power supplies? Greetings from Turkey.
Hi Dave I'm getting back into electronics after at least 30 yrs. I can't begin to tell you how useful and informative this and the follow up videos have been. thank you so much please please do more of this type of design videos
I first watched this video maybe 2 or 3 months ago and I had no idea what was going on because I had never done electronics before. Now I can see this video and follow it (at least). I think I have to return some more times till everything gets clear and I can say that I (will) have built my PSU and not only used some things from "the guy on youtube". Thanks Dave, I love you videos and hope you'll do them for a very long time!
you could just said: hey guys copy this schematic!, but your "ladder" explanation through all of the options from simpler to harder is brilliant. It's like university for us non engineers.. congrats!!
@joshstube Look at the LT3080 datasheet, you can parallel up devices to easily increase the current capability. It's 1A because that's all I wanted for my supply design. There are a zillion high current high voltage lab power supplies out there, this isn't going to be one of them. I may consider increasing to 2A by adding another device though. All will become clear in time.
Very nice progression from simplistic to more complex implementation of the design. It would be great to see an additional session dedicated to putting a "remote sense" in place...
You make it seem easy. Well, it's not that hard, but I feel rather motivated now. I'll just get a larger apartment so I can fit an electronics lab somewhere. Thanks, as always!
Only just watched this video and really enjoyed it Dave. I've always been fascinated by test equipment and power supplies especially. I've designed switched modes for a living but never got round to designing a decent bench power supply for myself. this might be the inspiration I need, off to watch part 2!!
there are a few reason for a constant current mode: Reason one is fail safe: The current limit protects both power supply and load from over current. You could design the supply to switch of hard, but... then you get an reset button which you have to press and a small peak in current (-> start up current) triggers the protection etc. The other application is current loads. Typical examples: (Power-) Leds, motors (when the should supply constant torque), magnetic actors etc.
I once built a switched mode power supply. Didn't work too well. I'm still proud of it, though. Especially since it was made almost exclusively from parts recycled from the ballasts of burnt out CFLs (being careful of course not to break the glass envelopes) The only parts I had to actually buy were a pnp transistor and a zener diode of the desired voltage. It doesn't have a pcb. Just soldered wires between all the parts and insulated it with duct tape. Good fun though.
@mitpatterson Oh ok, fair enough. That's not the way I typically use breadboards for simple stuff like this. Both top strips are power, and both bottom strips are ground. I think that's less confusing than having both ground and power on the top and bottom. It better follows how you would draw the schematic. Depends on the complexity of the circuit though, more complex builds can benefit from the other approach.
@Turkeylegs39 Have considered splitting into channels, but it just dilutes the content and the audience. I can't think of any benefits that outweigh the downsides.
Could you make a follow up video but with a switching power supply, buck/boost converter. Also it would be interesting to see a dual rail power supply.
For an easy to learn controller programmed in "Basic" that does have a ready to use DAC and code snippets try the "PICAXE" read the 3 manuals here. Grate video!
Dave; in short fantastic. Long form? The iterative design process - setting design parameters first - then engineering to meet those goals makes the 'math' problems far more understandable. Someone else has already suggested a 'tutorial' channel which is probably more work - but I try to watch every single tutorial I can. Thanks for you efforts.
Bryan Wenzel Considering fresh alkaline batteries are around 1.65V each, that's 6.6V nominal. Most likely, an extra 0.6V shouldn't damage the device. However, you can use a diode in series to drop 0.6V, or two in series to drop 1.2V if you are afraid of damaging the device. You may want a series resistor to limit the current of the lipo so it doesn't catch fire in case of a device fault.
filter capacitors. The remove ripples (one wanted fluctuations of the voltage) from the input and output voltages. They also serve as "energy storage" when the circuit needs for a very short amount of time more current. though due various effects (inductive, regulator speed etc.) the power supply circuit needs time before it can adapt to the higher current. In the mean time, the cap is supplying this peak current.
19:00 So I've gone and made that constant voltage circuit, and immediately had the exact issues you were talking about. I'm using a single NPN power transistor instead of a Darlington pair, but other than that it is effectively the same. My problem is that, while I can handle the logic and math of these things no problem, when it comes to dealing with the more complicated parts of dealing with characteristics of different voltages and resistances and crap that is where my knowledge falls apart.
You need to make one node of the NPN device lossy so it doesn't oscillate. For example you could insert a small value resistor (100-200 ohms) between the base of the NPN transistor and the output of the op-amp and also add some capacitance between the emitter of the NPN transistor and ground. By doing that you are actually modifying what is call the 'poles' of the feedback system. If you are interested look up stability of feedback systems.
This video is great, but the trapping is driving me nuts! Lol thank you for the amazing content, and thank you for teaching a person who is learning to solder, and build electronics on their own!
I am going to technical high school and I really must say that you explain much better than most of my teachers XD we are currently doing right that - internals of power regulators! I really love your vid's Now I see that I've missed whole "Lab power supply" series this is going to be a long night =D
@EEVblog @joshstube As I found out today in the Linear's magazine, there is now LT3083 which is good up to 3A, only downside is lower input/output voltage range. Price per one looks to be about double but I think that one would make decent PSU.
If i'm correcting myself, the output diode is to have a minimum load current. Also the diode will protect, if we accidently short the output to ground I guess.
u dont need a dac inside the micro for simple proyects. just use a pwm channel and use a low pass filter, if the sink and feed currents are simetric then the curve will be linear
In the first schematic he showed? Yup, the constant current pot would need to handle the full current. But think positive: It would reduce the thermal stress on both regulators.
I''ve built a two voltage regulator design ( had the parts at hand , and I didn't had any good opamps ) , but the current potentiometer overheats if set to more than 150mA . Any idea what can be tweaked to avoid that ?
why not use a buck-boost instead of linear? just a 555, 393, 385 and a couple of transistors and inductances from a scrap atx supply and a low resistance mosfet, and if you want it noise free add a lc filter it's not that hard and it's also easy to understand.
Very nice, Dave. BTW, I easily found the TI Data Sheet for this device and it has great info and application info that supports and supplements (dare I say compliments!) your friendly tutorial presentation with other benefits. Excellent! Cheers, mate.
Nice one! I was thinking to build a linear power supply with MCU control for a long time :) But to make the design with MOSFETS, jelly-bean op amps and use additional transformer tap for driving the mosfets and for simplified current sensing. And then I can implement some presets, OVP(limit), PC control and other stuff in software. But that's gonna be a major project.
Now I'll have to go back to my LM 317 DIY supply (ATX PSU based) and rebuild accordingly with the LT3080. Which means I'm going to have to learn more things. Damnit, Dave! :) Thanks for this.
+alecjahn If it works.....Don't "Fix" it...... (goes the saying) BTW you can accomplish this with an "Emitter-follower circuit" No Darlingtons, linear regs. etc. All you need is NPN power tabs. w a power tab the V at the emitter output will always be about .5 v Less than the base bc of the Semiconductor junction. Also remember that Transistors are CURRENT amplifiers, So.... All you need to know is the BETA of the Transistor and work w it for Current regulation........ and Last, this vblog dude has his brain scattered all over the place........
This idea can be easyli expanded to some sort of basic function generator ... Also consider some sort of PC connection just to add remote control or arbitrary wave form generator. Usign PWM looks nice.
@noob64ilive Yeah, DAC's are nicer for sure, but PWM is cheap and simple. A typical 10bit PWM in a micro is good enough for 10mV steps in a 10V supply.
man you're great. I really enjoyed this! Escpeccialy the time you took to explain the theory to us. I cant believe you re going to build it even if you know its not perfect yet xD
@33:30 Dave is telling us that we don't need the current-buffer anymore because the input to the opamp is "high impedance". How do I know that it is high impedance?
What if instead of using a small voltage reference like the 1.25V into a comparator for the negative feedback, and a voltage divider for the feedback, I instead had a reasonably high voltage reference or even a small voltage regulator connected to the 10 turn POT, or DAC and then the voltage output from that would be the an _adjustable_ reference, and the feedback was a permanent resistor divider (maybe a 25 turn trimmer for high accuracy calibration since I got a buttload of those.)? I should be able to get down into the millivolts or microvolts I would imagine! I designed a op-amp and 2N3055 based solution around this idea, and it works OK, until I try to implement current limiting. To do that, I have my 2N3055 transistor paired with a smaller 2N4011 as a darlington pair, and there is a 1Kohm resistor that pulls that base high, and then the 2 'comparators' (LM386 op amps) that also are tied to that base, but though some diodes so that either one can pull the base of the darlington config down low, without causing some parasitic current between the disagreeing voltage and current comparators. I basically ended up created a poor-mans OR gate, that works in the analog domain (lowest voltage on the op amp wins out and drives the base) The current comparator has a I ref, and on the inverting input, a differential amp (with the same op amp) connected to a shunt resistor, so that gives that current comparator has a gain of like 50 or so and the voltage output at the output is relative to ground rather than the high side shunt, which the voltage difference cannot be measured directly due to it being on the high side. I do have some issues with this in LTspice, Why is it that it likes to parasitically oscillate when I connect a small output capacitance, like 100nF to a few tens of uF. It seems backwards. I was able to fix it by adding a 100nF capacitor between the op amp output and the negative input, and a small 100 resistor between the input of the op amp and the voltage feedback, as well as for the current feedback. That seems to lower the instantaneous gain of the amp to zero, and over time the gain can integrate back up to nearly infinity, effectively becoming a low pass filter of sorts. Then the output is reasonably stable, but I am not sure how well it can cope with high frequency line rejection and load rejection. I was hoping the simulation was wrong, but building it up with junk parts on a breadboard showed the exact same things on my scope. Another issue is the drop voltage is really high. I need 4V more on the input than on the output for proper regulation. It seems to be mostly due to the darlington, but also the 0.1 ohm shunt at 5A. I considered using a P channel MOSFET but that seems even worse with parasitics. How do I get rid of these without excessive parts? Any suggestions?
What a fantastic day and age that we live in where generous individuals like you can teach through tutorials like this. Thank you so much for taking the time to do this, especially the theory explanation, I absolutely love it. Keep doing what you are doing!
Thanks, glad you found it useful.
EEVblog I am an almost complete newbie and looking through LDOs to build my own power supply (like real men do). I saw your new rigol PS and I figured I can code the software for one just like it, with color display, a nifty graycode encoder and all. I came across the same part and pretty much came up with the same schematic, except for the buffers which are an awesome fix I could not figure out (thank you for the opamp explanations as well. I finally understand the t-shirt). This tutorial is by far the best I've seen, for people like me, as it explains the train of thought that went into designing this which is far more important then etching techniques or explaining how to drill holes in aluminium. I also wanted to get more then 1 amp from it so what I did, since the part is based on a comparator, I put a PNP from vcc to output and it is opened by the _IN_ pin. Now I can get it up to 10 amps with proper cooling and a huge transistor. Spice shows some 20ma through the IC for a 3 amp output so I suppose I can get it up to 15-20 amps no problem, with an even bigger PNP and more cooling and it's still dirt cheap.
Well, generous... He was getting paid pretty well already by TH-cam when he shared this information. Don't get me wrong, I have nothing against David or this channel, I'm learning a lot from his videos too. But you have to be aware there's not just generosity at play here.
@@ArumesYT youtube doesnt pay very well, especially at the sort of range of veiws that he got at the time of your comment. his time could be more lucratively spent elswhere, so although he does get compensated for this, his decision to share this knowledge is definitely partially out of generosity.
@@ProteinFromTheSea So I say there's not JUST generosity at play. You counter that by saying it's at least PARTIALLY out of generosity. Why counter my comment when you agree? And Dave already was a full-time vlogger a this point, so apparently the pay was enough.
This is brilliant! You should do more of these longer "let's build something" series
I have 40 years of electrical engineering experience . Just out of curiosity I looked at your video. I must admit it is an excellent presentation. I liked very much the ingenious way you build the presentation with all the "twists" in the design.
Thank you very much for the brilliant presentation.
meirbns exactly what you did as an EE? 😃
you posted this in 2011 and still this is like the best video that explains how to make a power supply design ! thanks very much for this !
This is exactly what I've been looking for! All these years later and it's still relevant. Thanks!
i really can't express how helpful and useful this tutorial series is! Thanks a million for describing in so much details, it's like a 2nd year uni stuff for someone who never did Electronics beyond college. Love it!
@scientist375 Sure I could drop a few minutes here and there if I used a script or went and re-shoot some parts to make them tighter, but it might get 40min down to say 30min at best. You can't magically compress 40 minutes worth of material into 5 or 10 minutes and keep the same content and explanation. I could make this video 1 minute long and say here's the schematic I prepared earlier, but what's the fun in that.
The next video will be 20-30 minutes of looking at datasheets and tweaking
I personally much prefer proper, long form videos like you tend to make. It allows you to explore as in depth about the subject matter as you are so inclined. I love that, tbh.
@@MarkMcDaniel Me too. Natural talking without a script. Scripts make things feel forced and unnatural. I hate scripted videos.
"If you don't have specs to work from, well, it's gonna be a dog's breakfast."
Where do you come up with these great expressions? I love it!
He's from down under. I have and aussie friend he says things like that all the time.
Dave, I tip my hat to you. I've been pulling my hair out trying to work out how this is done and why the adjustable voltage is so simple yet the current limiting looks so complicated, but you have really spelled it out in a very descriptive way. Thanks so much!
@zox012 Yeah, I know. It was supposed to be one episode and I shot most of it thinking that, but when I got to editing, as is often the case, I was surprised at the amount of footage I shot. It was far too long, and kinda made sense to split up into a couple of episodes. So it should be at least 3 parts now.
Looking forward to the rest of the tutorial. I like how you evolved the design from using more components to a little more elegant with fewer components.
@scientist375 And the Makezine videos are quite lacking in detailed explained content. If you like that, fine, but it's not the way I like to explain stuff. I chose to show you some of my thought processes in the design, and that takes time. You can't magically shorten it, there is only so much you can fit in 5-10 minutes.
That's why I love this channel, Dave never misses a thing!
@mitpatterson It's orange :-> That's just the colour of various length pre-cut jumper links I have. I am using red and green for the power rails if you look on the left side.
This is still the first result for
power supply circuit design
Excellent 😁
Brilliant. I had to watch over & over 'till I understood. I don't mind the length at all. Its good for us who are learning.
Good Job Dave.
Your accent/voice is actually the main reason I'm here, and maybe enthusiasm about electronics design. Really great blog, honestly. May I request that you do a review about low-cost (0-80$) dc adjustable (0~30V) laboratory power supplies? Greetings from Turkey.
Hi Dave I'm getting back into electronics after at least 30 yrs. I can't begin to tell you how useful and informative this and the follow up videos have been. thank you so much please please do more of this type of design videos
I first watched this video maybe 2 or 3 months ago and I had no idea what was going on because I had never done electronics before. Now I can see this video and follow it (at least). I think I have to return some more times till everything gets clear and I can say that I (will) have built my PSU and not only used some things from "the guy on youtube".
Thanks Dave, I love you videos and hope you'll do them for a very long time!
you could just said: hey guys copy this schematic!, but your "ladder" explanation through all of the options from simpler to harder is brilliant. It's like university for us non engineers..
congrats!!
@joshstube Look at the LT3080 datasheet, you can parallel up devices to easily increase the current capability. It's 1A because that's all I wanted for my supply design. There are a zillion high current high voltage lab power supplies out there, this isn't going to be one of them. I may consider increasing to 2A by adding another device though.
All will become clear in time.
My new favorite video blog - thanks for all your work on these Dave :)
Dave the coolest guy I ever know!!
Keep sharing your knowledge Dave, thanks a lot!
Available in my Mech store. I drew it myself.
Very nice progression from simplistic to more complex implementation of the design. It would be great to see an additional session dedicated to putting a "remote sense" in place...
I appreciate all the help you have given me on my path to engineering. You are a great teacher.
@sorin0306 No, it won't affect the stability, the topology used ensures that.
You make it seem easy. Well, it's not that hard, but I feel rather motivated now. I'll just get a larger apartment so I can fit an electronics lab somewhere.
Thanks, as always!
Only just watched this video and really enjoyed it Dave. I've always been fascinated by test equipment and power supplies especially. I've designed switched modes for a living but never got round to designing a decent bench power supply for myself. this might be the inspiration I need, off to watch part 2!!
there are a few reason for a constant current mode: Reason one is fail safe: The current limit protects both power supply and load from over current. You could design the supply to switch of hard, but... then you get an reset button which you have to press and a small peak in current (-> start up current) triggers the protection etc.
The other application is current loads. Typical examples: (Power-) Leds, motors (when the should supply constant torque), magnetic actors etc.
I once built a switched mode power supply. Didn't work too well. I'm still proud of it, though. Especially since it was made almost exclusively from parts recycled from the ballasts of burnt out CFLs (being careful of course not to break the glass envelopes) The only parts I had to actually buy were a pnp transistor and a zener diode of the desired voltage. It doesn't have a pcb. Just soldered wires between all the parts and insulated it with duct tape. Good fun though.
@mitpatterson Oh ok, fair enough. That's not the way I typically use breadboards for simple stuff like this. Both top strips are power, and both bottom strips are ground. I think that's less confusing than having both ground and power on the top and bottom. It better follows how you would draw the schematic. Depends on the complexity of the circuit though, more complex builds can benefit from the other approach.
No, it's my natural voice.
THIS IS WONDERFUL WATCHING AND LISTENING TO THE LECTURE SESSION. MY LANGUAGE IS SPOKEN HERE. THANKS.
thanks to dave i finally understood how to set current with any micro controller. thank you so much dave!!!
@Turkeylegs39 Have considered splitting into channels, but it just dilutes the content and the audience. I can't think of any benefits that outweigh the downsides.
Could you make a follow up video but with a switching power supply, buck/boost converter. Also it would be interesting to see a dual rail power supply.
For an easy to learn controller programmed in "Basic" that does have a ready to use DAC and code snippets try the "PICAXE" read the 3 manuals here.
Grate video!
Thank you very much for the work your're doing for all of us. You guides are great!
"It's going to be a dog's breakfast..." LOL we never hear that in Minnesota.
@SajjadBro You mean actually getting a PWM signal out of a microcontroller? There are plenty of tutorials out there on how to do that.
Thanks you, this was one of your most interesting videos yet, taught me a lot!
Dave; in short fantastic. Long form? The iterative design process - setting design parameters first - then engineering to meet those goals makes the 'math' problems far more understandable. Someone else has already suggested a 'tutorial' channel which is probably more work - but I try to watch every single tutorial I can. Thanks for you efforts.
Bryan Wenzel Considering fresh alkaline batteries are around 1.65V each, that's 6.6V nominal. Most likely, an extra 0.6V shouldn't damage the device. However, you can use a diode in series to drop 0.6V, or two in series to drop 1.2V if you are afraid of damaging the device. You may want a series resistor to limit the current of the lipo so it doesn't catch fire in case of a device fault.
You're breathtaking.
After all these years
I heard Keanu reeves say that lol
Always A Pleasure to watch your Videos!!! Dave,, I Learn Learn Learn!!!! Thanks
Tank you Dave! Highly educational video for a second year electrical engineering student such as I
even i understood it omg, greetings from argentina thank you so much for the free classes.
filter capacitors. The remove ripples (one wanted fluctuations of the voltage) from the input and output voltages. They also serve as "energy storage" when the circuit needs for a very short amount of time more current. though due various effects (inductive, regulator speed etc.) the power supply circuit needs time before it can adapt to the higher current. In the mean time, the cap is supplying this peak current.
Click on the link below to a Digital power supply without a microcontroller:
A Digital Power Supply Without a Microcontroller!
Great video Dave!
best channel on youtube.
19:00 So I've gone and made that constant voltage circuit, and immediately had the exact issues you were talking about. I'm using a single NPN power transistor instead of a Darlington pair, but other than that it is effectively the same.
My problem is that, while I can handle the logic and math of these things no problem, when it comes to dealing with the more complicated parts of dealing with characteristics of different voltages and resistances and crap that is where my knowledge falls apart.
You need to make one node of the NPN device lossy so it doesn't oscillate. For example you could insert a small value resistor (100-200 ohms) between the base of the NPN transistor and the output of the op-amp and also add some capacitance between the emitter of the NPN transistor and ground. By doing that you are actually modifying what is call the 'poles' of the feedback system. If you are interested look up stability of feedback systems.
Oh man, I need more studying before I'm going to understand any of this.
4mb127 need more doing...
This video is great, but the trapping is driving me nuts! Lol thank you for the amazing content, and thank you for teaching a person who is learning to solder, and build electronics on their own!
I am going to technical high school and I really must say that you explain much better than most of my teachers XD
we are currently doing right that - internals of power regulators!
I really love your vid's
Now I see that I've missed whole "Lab power supply" series
this is going to be a long night =D
4:22 i think you can use single 317 for both current and Voltage controll
Thank you Mr Dave. It is a very good tutorial. I learned a lot. Now I am off to part 2.
@fridgebulb You need to follow me on Twitter :-> I slammed it in a door jam. Why does everyone assume I hit it with a hammer?
Good Stuff, just built a similar supply a little while back. Thanks for sharing. I've always enjoyed your videos.
@EEVblog @joshstube As I found out today in the Linear's magazine, there is now LT3083 which is good up to 3A, only downside is lower input/output voltage range. Price per one looks to be about double but I think that one would make decent PSU.
ah i hate it when good stuff is being presented and then "to be continued" :) really looking forward to part x
wow, I finally understood what that voltage inserted at the bottom via the voltage buffer does :) 22:51
If i'm correcting myself, the output diode is to have a minimum load current. Also the diode will protect, if we accidently short the output to ground I guess.
I love your 555 T-shirt
This is just brilliant.. I love the your way of starting simple.. best tutorial...
Thanks very much for this circuit, this is one of most important circuits for me
u dont need a dac inside the micro for simple proyects. just use a pwm channel and use a low pass filter, if the sink and feed currents are simetric then the curve will be linear
Some ideas:
1. Take a pot with higher current rating.
2. use a range switch and discrete power resistors (and a pot for fine adjustment)
In the first schematic he showed? Yup, the constant current pot would need to handle the full current. But think positive: It would reduce the thermal stress on both regulators.
Very nice demonstration.
Dave for president!
I''ve built a two voltage regulator design ( had the parts at hand , and I didn't had any good opamps ) , but the current potentiometer overheats if set to more than 150mA . Any idea what can be tweaked to avoid that ?
Really nice video, especially at 17:31!
Awesome first part, look forward to the follow up finishing it off!
One question though, whats up with your right thumb dave?
Thanks Dave Your explanation is fabulous
I learn a lot from you, i appreciate your videos. you have a massive amount of talent. keep up the good work. thank you.
Stm32f1 With integrated DAC only 2 bucks. Love those!
why not use a buck-boost instead of linear? just a 555, 393, 385 and a couple of transistors and inductances from a scrap atx supply and a low resistance mosfet, and if you want it noise free add a lc filter
it's not that hard and it's also easy to understand.
Very nice, Dave. BTW, I easily found the TI Data Sheet for this device and it has great info and application info that supports and supplements (dare I say compliments!) your friendly tutorial presentation with other benefits. Excellent! Cheers, mate.
Nice one! I was thinking to build a linear power supply with MCU control for a long time :) But to make the design with MOSFETS, jelly-bean op amps and use additional transformer tap for driving the mosfets and for simplified current sensing. And then I can implement some presets, OVP(limit), PC control and other stuff in software. But that's gonna be a major project.
eating pizza, chilling to your videos, is there anything better for an electronic engineer ?
Dave is a genius.
@Vlakpage Didn't know about that, thanks. Yeah, a bit pricey!
Now I'll have to go back to my LM 317 DIY supply (ATX PSU based) and rebuild accordingly with the LT3080. Which means I'm going to have to learn more things. Damnit, Dave! :)
Thanks for this.
+alecjahn If it works.....Don't "Fix" it...... (goes the saying)
BTW you can accomplish this with an "Emitter-follower circuit"
No Darlingtons, linear regs. etc.
All you need is NPN power tabs. w a power tab the V at the emitter output will always be about .5 v Less than the base bc of the Semiconductor junction.
Also remember that Transistors are CURRENT amplifiers, So.... All you need to know is the BETA of the Transistor and work w it for Current regulation........
and Last, this vblog dude has his brain scattered all over the place........
+Jess Cast Engineers version is: If it ain't broke, fix it!
hifatpeople Ummmm. I am an Engineer and I don't believe that is correct.
Jess Cast Just because something works, doesn't mean it can't be improved.
hifatpeople
That's true! It can also be fucked up and wasted.
This idea can be easyli expanded to some sort of basic function generator ...
Also consider some sort of PC connection just to add remote control or arbitrary wave form generator. Usign PWM looks nice.
Was that a motorcycle passing by at 1:10??
Anyway, thanks a lot Dave for all those many many hours of very informative videos!!
Nice design
2 cascade LM with variable R
One for constant current while the other to get constant voltage
This video was better than most of my electrical engineering lectures I paid thousands of dollars for
@EEVblog I was refering to the colors preprinted on the breadboard
@noob64ilive Yeah, DAC's are nicer for sure, but PWM is cheap and simple. A typical 10bit PWM in a micro is good enough for 10mV steps in a 10V supply.
Fantastic video! Thank you, Dave
Dave, I love you man... but please, I beg you... *please stop tapping the whiteboard with the damn pen!!!*
Overlord, I think I love you too, I was about to put the same message in, quickly read the comments and hey, you worded it perfectly!
I was not the only one that was annoyed with the pen 😂
Calm down, guys. It's not like nails across a chalkboard.
man you're great. I really enjoyed this! Escpeccialy the time you took to explain the theory to us. I cant believe you re going to build it even if you know its not perfect yet xD
If you enjoy his videos and watch them for longer than 2 days, you forget that.
dave as allways spot on!!!!!!!!!!!!!
@33:30 Dave is telling us that we don't need the current-buffer anymore because the input to the opamp is "high impedance". How do I know that it is high impedance?
Can you put the feedback loop on the LM317's adjust pin as in this design with the LT3080 to cancel out the 1.25v?
Why would you want to go so low on your regulator? Is there a component that uses less than 3 Volts?
that 1989 LM317 come back to work after 33 years...
What if instead of using a small voltage reference like the 1.25V into a comparator for the negative feedback, and a voltage divider for the feedback, I instead had a reasonably high voltage reference or even a small voltage regulator connected to the 10 turn POT, or DAC and then the voltage output from that would be the an _adjustable_ reference, and the feedback was a permanent resistor divider (maybe a 25 turn trimmer for high accuracy calibration since I got a buttload of those.)? I should be able to get down into the millivolts or microvolts I would imagine!
I designed a op-amp and 2N3055 based solution around this idea, and it works OK, until I try to implement current limiting. To do that, I have my 2N3055 transistor paired with a smaller 2N4011 as a darlington pair, and there is a 1Kohm resistor that pulls that base high, and then the 2 'comparators' (LM386 op amps) that also are tied to that base, but though some diodes so that either one can pull the base of the darlington config down low, without causing some parasitic current between the disagreeing voltage and current comparators. I basically ended up created a poor-mans OR gate, that works in the analog domain (lowest voltage on the op amp wins out and drives the base)
The current comparator has a I ref, and on the inverting input, a differential amp (with the same op amp) connected to a shunt resistor, so that gives that current comparator has a gain of like 50 or so and the voltage output at the output is relative to ground rather than the high side shunt, which the voltage difference cannot be measured directly due to it being on the high side.
I do have some issues with this in LTspice, Why is it that it likes to parasitically oscillate when I connect a small output capacitance, like 100nF to a few tens of uF. It seems backwards. I was able to fix it by adding a 100nF capacitor between the op amp output and the negative input, and a small 100 resistor between the input of the op amp and the voltage feedback, as well as for the current feedback. That seems to lower the instantaneous gain of the amp to zero, and over time the gain can integrate back up to nearly infinity, effectively becoming a low pass filter of sorts. Then the output is reasonably stable, but I am not sure how well it can cope with high frequency line rejection and load rejection. I was hoping the simulation was wrong, but building it up with junk parts on a breadboard showed the exact same things on my scope. Another issue is the drop voltage is really high. I need 4V more on the input than on the output for proper regulation. It seems to be mostly due to the darlington, but also the 0.1 ohm shunt at 5A. I considered using a P channel MOSFET but that seems even worse with parasitics. How do I get rid of these without excessive parts? Any suggestions?