I believe that regulator output has touched the heat sink which is directly connected to the collector of the big transistor which caused that Cap to blow
I'm surprised Sorin missed that. I see that a plastic insulator magically appeared under the replacement. I was waiting for the increasingly hot rheostat to be dropped onto the power supply.
I built that kit in 2019 - building it into a housing and added 10 turn pots and a Volt/amp meter. And it works great !! Best thing ...... unlike many of the constant current power modules you can buy - this one actually does a great job with current limiting !! It actually limits current down to 2.5 mA. Great for power injection and testing diodes !!
Same. I killed the first opamp from the output side with arking a street light ballast inductor. Perfect supply for the price. The input AC voltage is critical: the opamps use the full voltage rail with the -5V.
Gotta love the fact that your videos are so "organic". You also include fails along so we can learn from that, coming to a conclusion in the end and learning a new fix or a new way to blow up electronics, haha. Keep it up, Sorin! Looking forward to the next upload. Have a nice evening
12:40 if you go frame by frame first you see big spark that looks like it comes from transistor and then smoke + red glow from the capacitor on voltage regulator for the fan. I think that regulator touched fan pipe and capacitor on fan regulator shorted over the transistor 😎. FYI when you pause the video you can move forward and backwards frame by frame by comma and dot keys on the keyboard 🙂.
Learning by doing is the best way to lean and to achieve experience and not be afraid of the unknown or new. Smoke is good, smoking lessons from Sorin are funny, familiar friendly and very infesting always. All the best “moșu!”. Big hug!Cosmin from Switzerland 🇨🇭!
beware!!! :D that kit is wellknown, and have some flawless, too. the most easier flaw to fix is using for AO's a local power supply, like a Zenner and a resistor, calculated for no more than 30V over AO's, negative power included! it's a -5V on negative, so, 30-5=25, but there are no 25 V Zenners, not 3 terminal stabs, so, 24 volts. most easier is to use the already present 24V, for cooler. some small adjustment for resistors, and is perfect. of course, you need a nicely strong mains trafo, 24V or a bit more. or a bit less, if you accept a bit smaller voltage on full load... :) (written immediately after i seen the kit... is a wellknown for me! :D now, play to the finish... i'm really curious what you will find about... as a side note, is a really antique schematic, older than 60 years, in some variants. one of them use 3/4 AO's in a 324, and the last one is used as overload indicator, doubled as relay activator... :P ) later: min 13: to much current on a diode in the bridge! it's what happens most of time, without dealing with small smps's :D also, looks like you don't set the 0 volts, that blue multiturn is used to set the 0 volts, because this is a "virtual ground" power supply. for same reason you fried it first time with a 19 v laptop brick! :) lemme see... you just repeated the very same mistakes me and a lot of people does for the first time they meet this schematic, and, later, this kit... chinese started to manufacture and sell those kits some 15-18 years ago! :) i patiently seen the video to the final... Sorin, that blue led is NOT "power on" led, but "overload led", it need to be on only when you load the PS with more than your current limit! :) why is all the time on? some of these kits are correct, some not... some have a slight different schematic than the original in that part, and led is on the opposite directions, and is on all the time. other kits have wrong orientation printed on pcb. looks like electronics-diy point com shunned out that power supply, even this was their top-grade reason to check their website, a lot of years... you can find a lot with a google search, but i have trust in "radioscamateurs" sites, so... check the link down. i have not patience to search exactly about setting-up procedure, i had done it for some kits, and leaved it alone... also, with a bit of more searching, you can find a giga-talk about that schematic on a romanian forum, in power supplies section. a forum i leaved it for some years, even if i was one of the only posters there, for a lot of years, too... :) să fii iubit. :) www.qsl.net/z33t/dc_0-30v_0-3A_eng.html
13:35 - Strange thing you didnt stop video recording and check what sparked. Its power supply for fan, little capacitor. 12:39 spark. I believe that green pcb touched heatsink, and heatsink is soldered to the back of that big transistor. 18:29 you placed isolation tape on back :D
I noticed that your current limit led is constantly on, it should only come on when the current limiting kicks in. On my board the silkscreen had a error and the transistor that controls the led needs to be rotated 180 degrees that will make the led work correctly.
Fantastic, Sorin. My linear power supply i made with atx desktop power supply, using - cable blue -12v and yellow cable + 12v with aliexpress regulator 2956. -12v with +12v i have + 24v alright? Also, have a generic aliexpress meter voltage/current (als288 maybe) conected in serial mode with out - and out + . I am using 1 year no problem, but, i recomend using diodo retificator 4amp in +12 (its possible calibrate diference 0.7 volts on meter adjust) Its my homenade power supply setup; DIY Ever!
Definitely enjoyed this one, Sorin. One of your best. Please do more like this. It was funny with the flashback to the original barbecue video/livestream, too!
Nice! I build several working great though...but take care and replace IC with good ones, I'm using TL701 instead ''originals'', also I replaced R7 with 3x1ohm/5W in parallel , gives like 0.33 and let more current flow too.
did it ever work with the original op amps at all? Mine never worked with the originals, nor with the replacements I used (not a FET op amp, but no difference in behaviour)
@@AnnaVannieuwenhuyse my worked fine with originals but I somewhere read that they're very sensitive to overvoltage so I probly order replacement at same time. But PSU end up lost in mail for half a year so I in the end never really used it, just assembled and tested.
i built two of the power supplies powered up and op amps burned up so i change op-amp to lm741 because higher voltage rail and use dip sockets. i also also use 2n3055 as output transistor. used if for five years now
This guy is adorable! I like watching his videos, but this one is hilarious too. Kind of ElectroBoom. Depending on what you need the PSU for, I would consider buying 60$ labor PSU for repair and diagnostics.
Yes, usually is more comfortable to work with a 60$ PSU, but for some applications like developing an audio device with minimum requirements of noise linear PSU are the way to go.
As built the transistors are passing 5+ Amps (100+ Watts), but they are not dissipating anything near to that, which is why the fan/heatsink are coping and the transistors remain fairly cool. If you hooked up a 35V transformer as you suggested, then particularly when used at lower voltages the transistors will be dissipating much more power and thus get hotter, to the extent that you'll probably need a bigger fan/heatsink combination. Which is why many power supplies switch transformer windings in and out dependant on the voltage required at the output.
Nice video! Just one comment, you said the heatsink/transistors need to dissipate 100watts. But I guess most of the 100 watts gets lost as heat in your resistive load. You could measure it.
Why measure anything when he had all numbers in from of him on that little screen? Almost 20Vx5A=100W. But yes, in a resistor, transistors only dissipate difference between input and output voltage (xA of course) so around 25W with that transformer
With this sort of supply you need transistors and heatsinking adequate to handle the full rated output current with the output short circuited, or nearly shorted. That means that probably something in the range 120 to 150% of the rated normal maximum output power is dissipated in the transistors (the dissipation is the product of the output current and the *unregulated* voltage ahead of the transistors - i.e. on the main filter capacitor).
Would love to see an updated version video with all the mods you suggested a case and some testing done to it using your oscilloscope on the output for example. Plus if any more mods you can think of... Thanks for sharing though!!
I originally posted something like this in response to someone's question, but it seems a lot of people miss the issue, so I'm posting again with some elaboration. This is intended for use as a bench power supply. With a bench supply you really need to be able to deliver full rated current into a short circuit or close to a short. You may not need to do this often, but the need does arise. Your power supply needs to be able to survive it. When the output current is set to maximum and the output is shorted, the power dissipation in the transistors will actually be more than 100% of the full rated output power of the supply. If the supply is rated for 5 A and 30 V, that means the transistors would dissipate 150 watts, *minimum.* If fact the dissipation will be even greater because the voltage across the transistors will be equal to the full unregulated voltage from the bridge rectifier and filter capacitor. For the 30 volt supply that might be 35 volts (or more). At 35 volts the transistors would dissipate 175 watts at 5 A. That is a lot of heat to get rid of! If you use the regulator for some specific application where the output voltage is not very low, the situation is different. With high powers it is important to be careful that the transistors used have adequate safe operating area. "Second breakdown" can limit the allowable steady-state power dissipation to something well below the nominal rating of the device. With low voltage this often isn't a problem, but it very definitely becomes a problem when you get up into tens of volts.
Love the video😊I have the same power supply. It arived from china with dead amplifiers, and after few day transistor and big rezistor died😅I replaced them by original ST tranzistor and royal ohm rezistor and it works without problem😊
Hi Sorin, is current limiting functional? That diode on the pcb should light up only when overloaded. I have made a double stabilized source from these modules. And I had to replace all the ICs on both. And many other changes besides. Try to short-circuit the source, with a set current of 3A, and then you will see what the temperature will be on the transistors. :)
I went thru this adventure a couple of years ago and I was unlucky at first. Transistor blew up almost instantly at 2 amps. Later I got scammed with fake Toshiba transistors on amazon. Ended up using a 13007 extracted from an old ATX psu and problem was solved. If you have a proper power transistor with proper heatsinking, the thing does it's job. Have in mind that with this circuit, if it overheats and shorts, it will send the full unregulated power to whatever you have connected at the output.
@@spimfurt I don't know of any specific mod for that right now, I haven't done my research online. But it is actually a project I have been really wanting to make. I think by using a comparator IC we could make some kind of overvoltage protection that triggers a relay if it goes much higher than the voltage set.
Yes! A lot of people don't seem to get that. If you short the output, and a good bench supply should be able to handle that, the transistors will dissipate power equal to the 5 ampere current limit times the *unregulated* voltage, so something greater than the full rated output power of the supply. If the output can deliver 30 V and 5 A, the transistors would dissipate probably something in the range of 160 to180 watts.
@@d614gakadoug9 This is the reason why I was asking. Besides, an even better power supply would have a transformer with multitap secondary and an automatic relay logic to select the input voltage to never have a voltage drop on the output transistors greater than a certain maximum, to limit the power dissipated by them.
29:00 You are wrong. Here the variable resistor has to deal with 100W, the transistors not. If you configure a low output voltage at max current, then the transistors has to deal with 100W. I guess they will desolder from heatsink and blow up. Lets try. 😂
@@electronicsrepairschool No. The transistors are passing 5 amperes, that's true. But they will not dissipate 100 watts before they at the same time also have 20 Volta across collector/emitter. If you crank up the voltage to maximum it only be the Vce (of a bipolar transistor) times the current that is dissipated in the transistor. But if you set the output voltage to 1 volt and the transistors is fed 21 volts, then they have to "remove" / burn up the voltage difference which would be 20 volts. And at 5 amperes that would equal 100w of heat there.
@@electronicsrepairschool No, because if you have 24 V input, 20V output and 5A current, over the transistors you will have 4V drop. That means 20W of power disipation. If you have 24V input, 4V output and 5A, then you have 20V over the transistors and they have to tranfsform 100W electrical power into heat. That would be a real test. 🙂
I understand your point, actually is very easy to calculate, but, when you design/test a power supply, you take in consideration the worst case, in this case hare are not relays to switch the transformer output for low vce
@@electronicsrepairschool So why you didn't make the test with 1V output voltage and 5A output to show how hot the transistors can be ? With 20V output voltage is nothing to them.
The voltage on the output rectifier and capacitor filter is 24*sqrt(2) = above max rating of 7812 And seems small switching regulator short to the heatsink
Mulțumesc, Sorin. I wasn't sure how I was supposed to connect the Ammeter to the supply. My attempt resulted in the D882 transistor shorting out because I tried powering the meters from the fan output. Tried your method and it works flawlessly!
I built one of these kits, but haven't had a chance to even check if it works, as I don't have a suitable transformer. it was an entertaining enough project, and maybe I'll find a decent 24v transformer in something at some point in the future to test it out.
i bought this a year ago. the achiles heel here is the low voltage and high current. a 3V motor powered by 1.4V and 0.8 - 1.5A for an hour caused the failure of the big transistor. the loud one. sorin, i will appreciate if you can find a way to make this power source to sustain 0.8V and 3A forewer. typicaly for ram/cpu power testing. for cooling i used the 775 heatsink from intel quad core cpu. no fan. transformers are available in some cheap speakers with amplifier. usually 2.5A, -11 and +11V = 22V. or this expensive sorins fancy toroid.
Quite hard but not impossible. In the situation you mention It Will dissipate more than 60w in heat. You Just use a bigger transistor and Active cooling
how about a hybrid powersupply? there the final regulation is done by a transistor operating in its linear region - but a switching mosfet pre regulator - that tracks the voltage to say - 3volts higher than the transistor to make the power that the transistor has to turn into heat that much less?
That can work quite well though dynamic performance may not be as good as with an all-linear design. You still need pass transistors rated for the maximum input voltage so that if the load is abruptly removed after running at full voltage (meaning the switcher is putting out its max voltage) the transient input-output differential across the pass device won't kill it. If the load is very abruptly increased it may take the switcher a bit of time to "catch up." With modern designs running at quite high frequency and carefully done for good dynamic performance this may not be an issue. I have watched quite a few videos of amateurs making power supplies. So far not a single one has done anything at all to test dynamic performance. Poor dynamics can lead to voltage drops which can cause circuit misoperation but usually aren't disastrous, or overshoots that can destroy circuitry being powered. Everybody just connects a meter and calls everthing good if the meter looks stable. Well, the typical DVM will show you 25 volts for a 5 kHz 100 V rectangular waveform with 25% duty cycle, so ...
on hight currents and low voltages, the powersupply is capable of one, but loud, click. but for real, that blue led goes off, when the short is present.
hello zorin, can you detailed the connection using the ac transformer to the board to avoid blowing up the 3 ic's, planning to follow these project of yours. thanks a lot.. a follow up video or explanation is so much appreciated. more power to your channel. :)
What I think happened is the bottom of the small DC-DC converter touched the copper heat pipe which is soldered to the collector of the big transistor, so.........you can figure out the rest :))
@@repairtech9320 The odd thing is the spark appears to originate from the transistor soldered to the heatsink, rather than the dc-dc converter. After the spark the capacitor on the dc-dc board glows red before it burns up, Not watched any further yet, so will see what else happens...
I still have this kit somewhere assembled, ending up never using it but is there as backup, just in case. I got it for free because it was lost in mail somewhere so aliexpress give me refund, package arrive 6 months after that. btw fan will have much more problem with lower voltage, with 19V transistors don't need to drop too much but if you lower your voltage to lets say 1V and try to push 5A for long time that little heatsink would have hard time :)
Salut Sorin. eu am facut aceasta sursa de alimentare si tot mi se ardea OP-ampurile si nu stiam de ce. eu am cumparat un traf toroidal de 24v AC dar in gol scotea 27V. traful trebuie sa scoata fix 24 V in gol ca sa nu arda integratele. 24V AC inmultit cu constanta capacitiva de 1.4 dupa redresare si filtrare o sa ai in jur de 33.6V maxim. cu 27V in gol inmultit cu 1.4 ne da 37.8V care ard integratele pentru ca nu sunt proiectate pentru o astfel de tensiune. tensiunea maxima admisa pentru TL081 este de 30V DC. ce este peste le arde si nu stiti de ce sursele nu functioneaza. pentru integrate folositi un soclu si a-ti scapat de problema cand ele se ard. frumoase videoclipuri Sorine. Ma bucur ca avem asa romani priceputi si peste hotare. o zi faina!
Nice video.keep going sorin,may i ask what model is the camera you record with the one you use to record from top,i checked you tools link but you haven't included
If you slow down the video, it literally looks like some voltage flies from the right corner of the screen and trashes your 5v out. I'm not sure what happened there, but it looked like something zapped it's tiny ceramic capacitor from several inches away on your work bench.
12:39 minutes I think you didn't know the maximum voltage input of your mini step down dc to dc converter, which is connected to your cooling fan , i saw that smd filter capacitor burn out, as I've zoom in the video, 😅 you mishandled it don't you, better luck nxt must check your module's capacity input before connecting into the input supply 😢
Hi sorin great video again I would love to see how you would build or modify a ebike motor controller I want to make one that can maybe run with 1 36v battery and be able to adjust or just boost the power to the motor for short periods or to run it at 72v I think it would be a good video especially using your special calibrated wires 😂 ect. Thanks for all the really helpful stuff you have taught me, hope you are well.
On mine the output to the display is up to 33v. Connected as you do. The display is rated 4.5 to 30v. I touched its components and the transistor gets very hot. So a step down regulator is needed?
hi Sorin, I'm using this circuit since 6 yrs. lately D882 get burnt, I replace it but agian burnt, I found the cap 0.22u next to D882 damaged I replaced with used one works for a while but again D882 get burn i remove the cap and it worked fine till now, couldn't figure why.
If I insert a 24volt Zener diode in the regulator ground lead this will in effect increase the max output to 48 volts which is what I need , however can the rest of the components take the increase in voltage?
The little dcdc regulators need good output caps (x5r or x7r) and Chinese wants more profit so they put junk caps. Simply check ripple with dmm set to ac mode.
Sorin be careful, warming up a heatpipe that much can very easily make it explode. And if it doesnt explode, it loses some of the ability to move heat to the heatsink
Hello. The problem was the short that was mistakenly made when regulator board touch the heatsink pipe, which is from copper. One must be more cautious.
Hello, interesting video. Can you share the parts to understand how it works (Transistors, chips...) Thanks. BTW, can we replace those transistors by N Channel mosfets ? (more available and can take more current)
Great switching mosfets are not necessarily great in a linear supply. I would guess that the transistor package and heat sink (and the voltage difference/power dissipated) determine how much current it can take.
I believe that regulator output has touched the heat sink which is directly connected to the collector of the big transistor which caused that Cap to blow
Yep.
yes - i think thats what happened too
if it was 230V regulator he would notice it 😂
@@nicmate2543
absolutely 😄
I'm surprised Sorin missed that. I see that a plastic insulator magically appeared under the replacement. I was waiting for the increasingly hot rheostat to be dropped onto the power supply.
I built that kit in 2019 - building it into a housing and added 10 turn pots and a Volt/amp meter. And it works great !! Best thing ...... unlike many of the constant current power modules you can buy - this one actually does a great job with current limiting !! It actually limits current down to 2.5 mA. Great for power injection and testing diodes !!
Same. I killed the first opamp from the output side with arking a street light ballast inductor. Perfect supply for the price. The input AC voltage is critical: the opamps use the full voltage rail with the -5V.
Gotta love the fact that your videos are so "organic". You also include fails along so we can learn from that, coming to a conclusion in the end and learning a new fix or a new way to blow up electronics, haha. Keep it up, Sorin! Looking forward to the next upload. Have a nice evening
12:40 if you go frame by frame first you see big spark that looks like it comes from transistor and then smoke + red glow from the capacitor on voltage regulator for the fan. I think that regulator touched fan pipe and capacitor on fan regulator shorted over the transistor 😎.
FYI when you pause the video you can move forward and backwards frame by frame by comma and dot keys on the keyboard 🙂.
Great tip on using the comma & period for frame by frame. Thanks!
The best way to learn electronics is this building something from zero. Nice one
It's funny... you went from "Electronics Repair School" into "ElectroBoom" in this video... loved it Sorin... :)
Learning by doing is the best way to lean and to achieve experience and not be afraid of the unknown or new. Smoke is good, smoking lessons from Sorin are funny, familiar friendly and very infesting always. All the best “moșu!”. Big hug!Cosmin from Switzerland 🇨🇭!
beware!!! :D
that kit is wellknown, and have some flawless, too.
the most easier flaw to fix is using for AO's a local power supply, like a Zenner and a resistor, calculated for no more than 30V over AO's, negative power included!
it's a -5V on negative, so, 30-5=25, but there are no 25 V Zenners, not 3 terminal stabs, so, 24 volts.
most easier is to use the already present 24V, for cooler.
some small adjustment for resistors, and is perfect.
of course, you need a nicely strong mains trafo, 24V or a bit more. or a bit less, if you accept a bit smaller voltage on full load... :)
(written immediately after i seen the kit... is a wellknown for me! :D now, play to the finish... i'm really curious what you will find about... as a side note, is a really antique schematic, older than 60 years, in some variants. one of them use 3/4 AO's in a 324, and the last one is used as overload indicator, doubled as relay activator... :P )
later: min 13: to much current on a diode in the bridge! it's what happens most of time, without dealing with small smps's :D
also, looks like you don't set the 0 volts, that blue multiturn is used to set the 0 volts, because this is a "virtual ground" power supply. for same reason you fried it first time with a 19 v laptop brick! :)
lemme see... you just repeated the very same mistakes me and a lot of people does for the first time they meet this schematic, and, later, this kit... chinese started to manufacture and sell those kits some 15-18 years ago! :)
i patiently seen the video to the final... Sorin, that blue led is NOT "power on" led, but "overload led", it need to be on only when you load the PS with more than your current limit! :)
why is all the time on? some of these kits are correct, some not... some have a slight different schematic than the original in that part, and led is on the opposite directions, and is on all the time. other kits have wrong orientation printed on pcb.
looks like electronics-diy point com shunned out that power supply, even this was their top-grade reason to check their website, a lot of years...
you can find a lot with a google search, but i have trust in "radioscamateurs" sites, so... check the link down. i have not patience to search exactly about setting-up procedure, i had done it for some kits, and leaved it alone...
also, with a bit of more searching, you can find a giga-talk about that schematic on a romanian forum, in power supplies section. a forum i leaved it for some years, even if i was one of the only posters there, for a lot of years, too... :)
să fii iubit. :)
www.qsl.net/z33t/dc_0-30v_0-3A_eng.html
Do you think is possible to kill this kit by connecting a cara battery to It with out the Power supply be set to a voltage above the battery voltage?
The little buck converter board was sitting right on top of the copper heat sink. Wild!
12:39 You shorted the little switching power supply with (heatpipe cooler) the collector of the transistor
Thank you, your videos make electronics fun again.
so satisfying to watch the soldering process
13:35 - Strange thing you didnt stop video recording and check what sparked. Its power supply for fan, little capacitor. 12:39 spark.
I believe that green pcb touched heatsink, and heatsink is soldered to the back of that big transistor. 18:29 you placed isolation tape on back :D
It's not only you, when I try to build something to show someone, sometimes I have such "luck" as well.
I noticed that your current limit led is constantly on, it should only come on when the current limiting kicks in. On my board the silkscreen had a error and the transistor that controls the led needs to be rotated 180 degrees that will make the led work correctly.
Fantastic, Sorin. My linear power supply i made with atx desktop power supply, using - cable blue -12v and yellow cable + 12v with aliexpress regulator 2956. -12v with +12v i have + 24v alright? Also, have a generic aliexpress meter voltage/current (als288 maybe) conected in serial mode with out - and out + . I am using 1 year no problem, but, i recomend using diodo retificator 4amp in +12 (its possible calibrate diference 0.7 volts on meter adjust) Its my homenade power supply setup; DIY Ever!
Definitely enjoyed this one, Sorin. One of your best. Please do more like this. It was funny with the flashback to the original barbecue video/livestream, too!
That regulator for the fan made contact with the copper from the cooler and got shorted because you didn't isolated the board under! 🤣🤣🤣
Great tutorial! Nice music at 23:55 😉
Nice! I build several working great though...but take care and replace IC with good ones, I'm using TL701 instead ''originals'', also I replaced R7 with 3x1ohm/5W in parallel , gives like 0.33 and let more current flow too.
did it ever work with the original op amps at all? Mine never worked with the originals, nor with the replacements I used (not a FET op amp, but no difference in behaviour)
now I finally know for what I order bunch TL701 three years ago :D
@@AnnaVannieuwenhuyse my worked fine with originals but I somewhere read that they're very sensitive to overvoltage so I probly order replacement at same time. But PSU end up lost in mail for half a year so I in the end never really used it, just assembled and tested.
do u know where i can buy transformer? 20v 5A
@@VanDuc1126 I built mine th-cam.com/video/yxEW2CWy4ew/w-d-xo.html
there is that properly calibrated wire love it man thanks for video been trying to find parts for making one from junk tvs
Ive been expecting this so long thanks Sorin
Belle vidéo, j'ai aussi réalisé cette alimentation, j'ai appris des choses grâce à vous, merci !
i built two of the power supplies powered up and op amps burned up so i change op-amp to lm741 because higher voltage rail and use dip sockets. i also also use 2n3055 as output transistor. used if for five years now
This guy is adorable! I like watching his videos, but this one is hilarious too. Kind of ElectroBoom.
Depending on what you need the PSU for, I would consider buying 60$ labor PSU for repair and diagnostics.
Yes, usually is more comfortable to work with a 60$ PSU, but for some applications like developing an audio device with minimum requirements of noise linear PSU are the way to go.
As built the transistors are passing 5+ Amps (100+ Watts), but they are not dissipating anything near to that, which is why the fan/heatsink are coping and the transistors remain fairly cool. If you hooked up a 35V transformer as you suggested, then particularly when used at lower voltages the transistors will be dissipating much more power and thus get hotter, to the extent that you'll probably need a bigger fan/heatsink combination. Which is why many power supplies switch transformer windings in and out dependant on the voltage required at the output.
Nice video! Just one comment, you said the heatsink/transistors need to dissipate 100watts. But I guess most of the 100 watts gets lost as heat in your resistive load. You could measure it.
Why measure anything when he had all numbers in from of him on that little screen? Almost 20Vx5A=100W. But yes, in a resistor, transistors only dissipate difference between input and output voltage (xA of course) so around 25W with that transformer
Correct . A piece of copper wire carrying 5amps at 20volts does not dissipate 100watts!..it’s 5amps x the voltage drop of the wire !
With this sort of supply you need transistors and heatsinking adequate to handle the full rated output current with the output short circuited, or nearly shorted. That means that probably something in the range 120 to 150% of the rated normal maximum output power is dissipated in the transistors (the dissipation is the product of the output current and the *unregulated* voltage ahead of the transistors - i.e. on the main filter capacitor).
Would love to see an updated version video with all the mods you suggested a case and some testing done to it using your oscilloscope on the output for example. Plus if any more mods you can think of... Thanks for sharing though!!
I originally posted something like this in response to someone's question, but it seems a lot of people miss the issue, so I'm posting again with some elaboration.
This is intended for use as a bench power supply. With a bench supply you really need to be able to deliver full rated current into a short circuit or close to a short. You may not need to do this often, but the need does arise.
Your power supply needs to be able to survive it.
When the output current is set to maximum and the output is shorted, the power dissipation in the transistors will actually be more than 100% of the full rated output power of the supply. If the supply is rated for 5 A and 30 V, that means the transistors would dissipate 150 watts, *minimum.* If fact the dissipation will be even greater because the voltage across the transistors will be equal to the full unregulated voltage from the bridge rectifier and filter capacitor. For the 30 volt supply that might be 35 volts (or more). At 35 volts the transistors would dissipate 175 watts at 5 A. That is a lot of heat to get rid of!
If you use the regulator for some specific application where the output voltage is not very low, the situation is different.
With high powers it is important to be careful that the transistors used have adequate safe operating area. "Second breakdown" can limit the allowable steady-state power dissipation to something well below the nominal rating of the device. With low voltage this often isn't a problem, but it very definitely becomes a problem when you get up into tens of volts.
I feel really good about my soldering skills after seeing this 😂 but seriously thank you for showing me this kit
Use your new oscilloscope on the outputs to see how noisy it is
It's a linear, regulated power supply. The noise will be very, very little, almost nonexistent.
@@Atelierul29 th-cam.com/users/shortszF4elc9cnzM
Love the video😊I have the same power supply. It arived from china with dead amplifiers, and after few day transistor and big rezistor died😅I replaced them by original ST tranzistor and royal ohm rezistor and it works without problem😊
Hi Sorin, is current limiting functional? That diode on the pcb should light up only when overloaded. I have made a double stabilized source from these modules. And I had to replace all the ICs on both. And many other changes besides. Try to short-circuit the source, with a set current of 3A, and then you will see what the temperature will be on the transistors. :)
You have right, only if in short that led will be light
As always Brilliant Just wished you pointed out how important the input toroid/transformer is.
12:38 Sorin, the power supply specialist! 🙂
I went thru this adventure a couple of years ago and I was unlucky at first. Transistor blew up almost instantly at 2 amps. Later I got scammed with fake Toshiba transistors on amazon. Ended up using a 13007 extracted from an old ATX psu and problem was solved. If you have a proper power transistor with proper heatsinking, the thing does it's job. Have in mind that with this circuit, if it overheats and shorts, it will send the full unregulated power to whatever you have connected at the output.
is there a mod to avoid this? have a similar experience. thanks for the 13007 tip.
@@spimfurt I don't know of any specific mod for that right now, I haven't done my research online. But it is actually a project I have been really wanting to make. I think by using a comparator IC we could make some kind of overvoltage protection that triggers a relay if it goes much higher than the voltage set.
Sorin is my favorite influencer in the repair industry. Hi really love what is he doing
23:01 Do the transistors bear 5A on a short circuit, when they have to dissipate all the power?
Yes!
A lot of people don't seem to get that. If you short the output, and a good bench supply should be able to handle that, the transistors will dissipate power equal to the 5 ampere current limit times the *unregulated* voltage, so something greater than the full rated output power of the supply. If the output can deliver 30 V and 5 A, the transistors would dissipate probably something in the range of 160 to180 watts.
@@d614gakadoug9 This is the reason why I was asking. Besides, an even better power supply would have a transformer with multitap secondary and an automatic relay logic to select the input voltage to never have a voltage drop on the output transistors greater than a certain maximum, to limit the power dissipated by them.
@@you2ber252
Switiching taps can result in poor dynamic performance. It is much more common in cheap supplies than high-end bench supplies.
@@d614gakadoug9 We're talking of hobbyist gear here
29:00 You are wrong. Here the variable resistor has to deal with 100W, the transistors not. If you configure a low output voltage at max current, then the transistors has to deal with 100W. I guess they will desolder from heatsink and blow up. Lets try. 😂
Who is delivering the 100W to the resistor? :D Actually based on the transistors efficiency, for 100W output, the transistors has to deal around 130w
@@electronicsrepairschool No. The transistors are passing 5 amperes, that's true. But they will not dissipate 100 watts before they at the same time also have 20 Volta across collector/emitter.
If you crank up the voltage to maximum it only be the Vce (of a bipolar transistor) times the current that is dissipated in the transistor. But if you set the output voltage to 1 volt and the transistors is fed 21 volts, then they have to "remove" / burn up the voltage difference which would be 20 volts. And at 5 amperes that would equal 100w of heat there.
@@electronicsrepairschool No, because if you have 24 V input, 20V output and 5A current, over the transistors you will have 4V drop. That means 20W of power disipation.
If you have 24V input, 4V output and 5A, then you have 20V over the transistors and they have to tranfsform 100W electrical power into heat. That would be a real test. 🙂
I understand your point, actually is very easy to calculate, but, when you design/test a power supply, you take in consideration the worst case, in this case hare are not relays to switch the transformer output for low vce
@@electronicsrepairschool So why you didn't make the test with 1V output voltage and 5A output to show how hot the transistors can be ? With 20V output voltage is nothing to them.
I bought few years ago one of the RD tech power supplies from aliexpress, I have a 32V 5A and it's been very good.
but it is not linear . it is switching and you have ripple
The voltage on the output rectifier and capacitor filter is 24*sqrt(2) = above max rating of 7812
And seems small switching regulator short to the heatsink
Maximum voltage for regulators between 5...18V is 35V, so 33,84V is under the maximum limit. Any datasheet will confirm that!
Mulțumesc, Sorin. I wasn't sure how I was supposed to connect the Ammeter to the supply. My attempt resulted in the D882 transistor shorting out because I tried powering the meters from the fan output. Tried your method and it works flawlessly!
I built one of these kits, but haven't had a chance to even check if it works, as I don't have a suitable transformer. it was an entertaining enough project, and maybe I'll find a decent 24v transformer in something at some point in the future to test it out.
Old 24V UPS is a great source.
Well, you can power it with DC without any modifications.
@@ivankirola2707 You can't, it won't be able to generate the negative rail it needs.
i bought this a year ago. the achiles heel here is the low voltage and high current. a 3V motor powered by 1.4V and 0.8 - 1.5A for an hour caused the failure of the big transistor. the loud one. sorin, i will appreciate if you can find a way to make this power source to sustain 0.8V and 3A forewer. typicaly for ram/cpu power testing. for cooling i used the 775 heatsink from intel quad core cpu. no fan. transformers are available in some cheap speakers with amplifier. usually 2.5A, -11 and +11V = 22V. or this expensive sorins fancy toroid.
Quite hard but not impossible. In the situation you mention It Will dissipate more than 60w in heat. You Just use a bigger transistor and Active cooling
really the case for any linear supply
how about a hybrid powersupply? there the final regulation is done by a transistor operating in its linear region - but a switching mosfet pre regulator - that tracks the voltage to say - 3volts higher than the transistor to make the power that the transistor has to turn into heat that much less?
Or multi step transformer, like 0-5-10-15-20-25-30V.
That can work quite well though dynamic performance may not be as good as with an all-linear design.
You still need pass transistors rated for the maximum input voltage so that if the load is abruptly removed after running at full voltage (meaning the switcher is putting out its max voltage) the transient input-output differential across the pass device won't kill it.
If the load is very abruptly increased it may take the switcher a bit of time to "catch up." With modern designs running at quite high frequency and carefully done for good dynamic performance this may not be an issue.
I have watched quite a few videos of amateurs making power supplies. So far not a single one has done anything at all to test dynamic performance. Poor dynamics can lead to voltage drops which can cause circuit misoperation but usually aren't disastrous, or overshoots that can destroy circuitry being powered. Everybody just connects a meter and calls everthing good if the meter looks stable. Well, the typical DVM will show you 25 volts for a 5 kHz 100 V rectangular waveform with 25% duty cycle, so ...
great video and great cheap supply i have it too 😉
How can we add the relay for clicking effect of short circuit?
Good question!
on hight currents and low voltages, the powersupply is capable of one, but loud, click. but for real, that blue led goes off, when the short is present.
hello zorin, can you detailed the connection using the ac transformer to the board to avoid blowing up the 3 ic's, planning to follow these project of yours. thanks a lot.. a follow up video or explanation is so much appreciated. more power to your channel. :)
What I think happened is the bottom of the small DC-DC converter touched the copper heat pipe which is soldered to the collector of the big transistor, so.........you can figure out the rest :))
in slow motion you can see the spark coming out from under the 5V PS shorting on the heat pipe
@@repairtech9320 The odd thing is the spark appears to originate from the transistor soldered to the heatsink, rather than the dc-dc converter. After the spark the capacitor on the dc-dc board glows red before it burns up, Not watched any further yet, so will see what else happens...
@@will_doherty true
I still have this kit somewhere assembled, ending up never using it but is there as backup, just in case. I got it for free because it was lost in mail somewhere so aliexpress give me refund, package arrive 6 months after that.
btw fan will have much more problem with lower voltage, with 19V transistors don't need to drop too much but if you lower your voltage to lets say 1V and try to push 5A for long time that little heatsink would have hard time :)
thank you very useful, greetings from Indonesia.
you need to fetch 2A output @2V to feel the hotness at the transistors.
Thank you Sorin.....Like this DIY theme
Haha, time flies. I remember it like it was yesterday.
Salut Sorin. eu am facut aceasta sursa de alimentare si tot mi se ardea OP-ampurile si nu stiam de ce. eu am cumparat un traf toroidal de 24v AC dar in gol scotea 27V. traful trebuie sa scoata fix 24 V in gol ca sa nu arda integratele. 24V AC inmultit cu constanta capacitiva de 1.4 dupa redresare si filtrare o sa ai in jur de 33.6V maxim. cu 27V in gol inmultit cu 1.4 ne da 37.8V care ard integratele pentru ca nu sunt proiectate pentru o astfel de tensiune. tensiunea maxima admisa pentru TL081 este de 30V DC. ce este peste le arde si nu stiti de ce sursele nu functioneaza. pentru integrate folositi un soclu si a-ti scapat de problema cand ele se ard. frumoase videoclipuri Sorine. Ma bucur ca avem asa romani priceputi si peste hotare. o zi faina!
No capacitor No problem love it 😊
Nice video.keep going sorin,may i ask what model is the camera you record with the one you use to record from top,i checked you tools link but you haven't included
When you put the variable resistor on the bench next to the toroid you got an induced current which was too much for the capacitor.
You are going too far away! He shorted it to the heatsink.
Properly calibrated fire extinguisher in order sorin. I use one myself. And remember, no capacitor, no shorted capacitor. 😂
Finaly good quality picture from microscope!!!
but Sorin. The transistiors are not actually consuming 100W, so cooling should be fine. most of the power is taken by the resistor/load
I have lot of d288 mosfet , i am searching project power supply for using this mosfets
If you slow down the video, it literally looks like some voltage flies from the right corner of the screen and trashes your 5v out. I'm not sure what happened there, but it looked like something zapped it's tiny ceramic capacitor from several inches away on your work bench.
Sorine uzivanje je te gledati💪💪💪👍👍👍🇷🇸❤️🇷🇴
Can you please show the quality of its signal on the new oscilloscope you bought?
on that lil 5v board, the cap on the end blew... magic smoke 🚬 .. i watched it frame by frame...
Thank you. great work.
12:39 minutes I think you didn't know the maximum voltage input of your mini step down dc to dc converter, which is connected to your cooling fan ,
i saw that smd filter capacitor burn out, as I've zoom in the video, 😅 you mishandled it don't you, better luck nxt must check your module's capacity input before connecting into the input supply 😢
Very nice sir..I'll try to make this..
Sorin, after hour from you published that video there is not in stock in Amazon. Your audience is extremely loyal
They are not sold out, but probably Amazon have restriction based on the zone/continent/country. Check aliexpress or search on your amazon/ebay
nu te supara dar am ras ca lumea la 12:40 ,... multumim mult pt video. foarte bun,....
Regulators dont need to dissipate the full power to the load i.e. 100W but a fraction of that, that's why your fan made it.
Hi sorin great video again I would love to see how you would build or modify a ebike motor controller I want to make one that can maybe run with 1 36v battery and be able to adjust or just boost the power to the motor for short periods or to run it at 72v I think it would be a good video especially using your special calibrated wires 😂 ect. Thanks for all the really helpful stuff you have taught me, hope you are well.
17:50 MTGA movement motto "Ok lets use proper calibrated wire to Make Track Great Again ..."
Hello Sorin. Will you be installing small value resistors in the emitters of the those 2 transistors?
Hello🤝very nice power supplies 👍👍👍👍👏👏👏👏👏👋👋👋
You've spoilt us today 😂😂 🔥🔥
Thanks for the info. 👍
No ICs were hurt in this video! Awesome
On mine the output to the display is up to 33v. Connected as you do. The display is rated 4.5 to 30v. I touched its components and the transistor gets very hot. So a step down regulator is needed?
Great Sorin.
hi Sorin, I'm using this circuit since 6 yrs. lately D882 get burnt, I replace it but agian burnt, I found the cap 0.22u next to D882 damaged I replaced with used one works for a while but again D882 get burn i remove the cap and it worked fine till now, couldn't figure why.
I still remember very well the First Video!
If I insert a 24volt Zener diode in the regulator ground lead this will in effect increase the max output to 48 volts which is what I need , however can the rest of the components take the increase in voltage?
Hello Sorin , i love how you fix things , i watched a lot of your video, may i know how can i buy this things by the way im from Philippines
Great video thank you.
Is it possible to use laptop mosfets to build a power supply. They have 30am ratings . Maybe some way to re use from old motherboards?.
thanks Sorin,
I bought the same kit, what kind/rating Transformer should i get for this kit to be able to get 30v 3A out of it?
Great video Sorin, I would use a bigger capacitor too on 5A for better filtering. Great stuff
i remember that bbq party lol , and i did it , thank you for all
The little dcdc regulators need good output caps (x5r or x7r) and Chinese wants more profit so they put junk caps. Simply check ripple with dmm set to ac mode.
I think I'll stick with the linear power supply I got from CPC! Great video as always though :)
Sorin be careful, warming up a heatpipe that much can very easily make it explode.
And if it doesnt explode, it loses some of the ability to move heat to the heatsink
need to lower the value of the current sense resistor and fit a higher wattage one too!
Whats the negative voltage doing in the curquit????
Hello. The problem was the short that was mistakenly made when regulator board touch the heatsink pipe, which is from copper. One must be more cautious.
I wonder if that kind of power supply can be used as an injection tool, I need one
Hello, interesting video. Can you share the parts to understand how it works (Transistors, chips...) Thanks.
BTW, can we replace those transistors by N Channel mosfets ? (more available and can take more current)
Great switching mosfets are not necessarily great in a linear supply. I would guess that the transistor package and heat sink (and the voltage difference/power dissipated) determine how much current it can take.
It be amazing if you make "upgrade" to the kit so the volt and amp can be adjusted slowly +- 0.1 v not jumping like the default
It just needs 10-turn pots to accomplish this. I have done this.
So i just need to change the included potentiometer with the 10 turn right?, and what are the pots value
@@raihanrizki1803 10k
@@raihanrizki1803 value isn't so important, you can go with same as original (10k if I recall correctly) but anything will work.
@@raihanrizki1803 they usually are 10K or 1K. Sometimes one 10K and the other 1K
the trace wants for 5A 110mil end emitter 0.1ohm resistor