12V 10A switching power supply (with schematic and explanation)

The schematic in my DB of reverse engineered schematics:
danyk.cz/reverz...
Today I made a teardown of an industrial switching power supply module S-120W-12 in a metal housing. The input is 100-240V AC mains and the output is 12V DC 10A 120W. It came from Ebay and it cost me only $8. In the listing it's called "switching power supply LED driver adapter", but it may have a lot of uses other than just powering LED strips. The SMPS is based on the old good UC3842 control chip, TL431 voltage reference and an 817 optocoupler. It has an open loop protection added, made of discrete transistors. It also has a softstart, a proper fuse, NTC thermistor, safety capacitors, snubber networks, EMI filters, good isolation distance and a relatively safe-looking transformer. The only problem of this switching power supply is the way the power transistor and diode are mounted on the heatsink! Just the sharp corner of it actually touches the aluminium plate! This video got quite long, but it's not easy to cover such a complex appliance in just few minutes. I didn't even get to explain why is the power supply clicking when in short circuit and why it may produce audible noise (whistling) if there was no slope compensation.
You can support me on Patreon:
/ diodegonewild
My Instagram:
/ savage_danyk

[@Rodderick45]

Excellent video on flyback switching power supply. I have 42 years of experience in the electrical and electronic field from power plant, consumer electronics, computer repair, aerospace CNC/NC maintenance, 17 years repairing and maintaining light rail commuter trains, attended 5 years of college of electrical engineering and numerous vocational and military schools. After saying that I want you to know you have provided answers to detail questions concerning switch mode power supplies which I have not been able to find. I have been working with switch mode power supplies for the last 36 years and have never seen this subject explained in detail and confidence the way you have done and now it’s all perfectly clear to me now from an design, engineering and repair point of view. The snubber network, continuous mode and discontinuous mode, the zener protection and the secondary regulation via 431 transistor using the voltage divider network is all perfectly clear to me now. Thank you very much for this video and keep putting out the videos. Knowledge and understanding of the buck, boost, flyback switchmode power supplies employing the principle of (Pwm) pulse width modulation is essential for today’s Technicians and engineers. Good job and I will be following you.

[@MlokKarel]

Dan really knows his sh*t, he would probably long be dead otherwise. Things he does with tesla coils are beyond crazy. But he really has a true deep understanding of how things work and what's still (reasonably:) safe and what's certainly better to avoid. And he's patient. And knows how to simplify things for explanation, yet not oversimplify them. Those are makings of all the great teachers. 👍

[@felixcat4346]

Great description of how a SMPS power supply work. No one on youtube even comes close.

[@DiodeGoneWild]

Thanks :)

[@omsingharjit]

Joseph Nicholas Exactly .

[@yktelectronics2411]

Joseph Nicholas Great!

[@SpencerSkelly]

Seriously

[@napoleon5125]

You should watch Jose caltum videos hi is Mexican speaking Spanish only

[@JoshVennix]

Who gives this video a dislike? Clearly a lot of time and effort has gone into reverse engineering and drawing up a schematic. Video like this are an excellent reference for repair and modification of these power supplies. Thanks mate. Keep up the good work!

[@rpbajb]

You have a lot of patience to reverse engineer this complex circuit. Thank you for all your work. I learned a lot.

[@Abc-qk1xt]

Engineer with a caaat...

[@SalimAhmad204682]

Very well explained. Not even my college lecturer come close during his lecture.

[@TajinderSingh-jl5wv]

Reverse tracing a circuit is something that not many people do today. And you explained it very well too. Thanks.

[@stephenphen4349]

Agree. So this things called as "reverse tracing circuit" or "reversed engineering schematic". I had done this experiment with THE REALLY SAME THING I've got from my college laboratory (SMPS S-100-12), and still I find it so damn HARD to reverse-scheme all the tracks when there are two sides in a PCB you should pay attention. it's REALLY TRICKY. and THIS MAN EXPLAINED IT ALL AS EASY AS HE SPEAK. He is an expert. Regards from Indonesia

[@stephenphen4349]

and today, when I tried to do this "reverse tracking circuit" activity, my friend told me that it's something foolish things to do because they said it is pointless to track a complicated circuit, as long as we know how to use the device, that's crazy. but i can't agree with them anymore, because here i found this man doing the same things as i did

[@omniyambot9876]

​@@stephenphen4349 Are you now experts?

[@sirajhussain791]

Excellent way of explanting better than University teachers, having understanding and clear concepts of Power Electronics, Electrical and Electronics.

[@philippe5394]

As rpbajb said "You have a lot of patience to reverse engineer the circuits", and then to explain us how it works and also you point the remarkable points. Thank you for all your work. I also learn a lot. A stroke to the cat...

[@edabrams5462]

I have several of these supplies that have been operating daily at about 80% load for about 4 years. I just had one fail so i took it apart and found 3 of the output caps had blown their tops off. Upon inspection I found they were 1000MFD 16V caps. Yes a 16V cap on a 12V power supply, I'm amazed it lasted this long! I replaced them with caps rated for 25V and were back in business. Thanks for an incredible explanation of how a SMPS works.

[@alexloktionoff6833]

Thank you for the video with the schematics and detailed explanation of work, also you've mentioned the working frequency. It was mandatory to adjust it, because all UC384x clones have a little different frequency formula. I was able to repair mine S-100-12 switching power supply!

[@PhillipLandmeier]

Very well done tear-down. I've designed switchers since about 1986 and am as shocked as you are about the poor mounting of the power semiconductors. On the other hand, I'm impressed that it didn't fail while you were testing it. After remounting those semiconductors properly to the heat sink, it's probably a solid little supply. Nice. Thanks for posting this.

[@SeanBZA]

Transistors also do double duty as short ciruit protection, as the feedback vanishes, but the loop will go to full power in an attempt to drive the output to get feedback. The transistors then reduce duty cycle to a low level that will not result in the power switch blowing up. Soft start circuit starts up initially as zero volts on the capacitor, and this is slowly charged up by the 5V supply through the resistor, the diode to feedback ensures that the lower of the capacitor voltage or the feedback transistor is the one controlling duty cycle, and when the capacitor charges up to 5V it is effectively out of circuit allowing the feedback to be the control. The poor mountring is typical, wrong lead dressing on the power devices ( leads cut short instead of being left full length) along with the wrong device locking tab, which is meant for a thicker power device, so is skew on the package and thus applies pressure to one side only. Best would be to shorten the leg slightly and put a silicone pad between the device and heatsink to improve thermal transfer. The 10A rating is likely the diode current rating, the long term rating would be 5A, and would be more so limited by the ripple current rating of the 3 output capacitors in the secondary, which probably only are rated for 1A5 of ripple current anyway. Probably will not have the longest life at 10A, likely to fail after around 2000H of operation, or 3 months of running full power 24/7, but running 8 hours a day might actually make a year before it goes bang as the output capacitors dry out and the control chip power supply capacitor goes high ESR. Fixed many of them with new capacitors on the output and those little 47uF capacitors on the primary side.

[@DiodeGoneWild]

Yes, the PNP transistors are also usefull to detect an overload beyond regulation. As one of the viewers noticed, the power diode (and maybe even the transistor) are recycled from an e-waste. The diode has a mark from a screw. This is why it couldn't have longer pins. They were already cut short when they salvaged it from an old device. You're right, the 47uF capacitors parallel to the chip power are very notoriuously going high ESR. Also the output capacitors won't last very long. A forward topology would be better for 120W, because most of the ripple current is filtered in an (usually toroidal iron dust cored) inductor before reaching the output capacitors.

[@omniyambot9876]

hello can you help me with uc3842 basics

[@AThreeDogNight]

Thank you for the great break down on how these work, I'm still learning electronics & you have gave the best description on these & on others that I've seen on you tube now in several months, you really know what your doing & it shows greatly. Again thank you for the help, Tom.

[@ibrahimteker2497]

This is the best expression of basic flyback circuitry with its all details. This is the good share for the beginners. Good job.

[@almnetx]

One of the best smps explanations. Just to add: nterference capacitors = mov. There are X movs and Y movs

[@FadiFadi00]

I only use this kind of power supply in my job.. which it’s work perfectly without any problems... the only problem is when it gets hot.. so cool it down and enjoy a high current, stable voltage, cheep power supply (: Nice description and explanation, i love your videos, it’s very informative and you care about safety the most!!!

[@cocosloan3748]

You explained this schematic very well-I learned a lot ! Thanks !!Greetings from Macedonia!

[@RegiPavan]

Great description of working of the power supply! And the cat is helping the class to relax...

[@rikka0_059]

tl431 here works like a opamp to form a type-2 loop compensator. It provides very high open loop dc gain so that the output is well regulated. The RC near tl431 form one zero and the 100nf capacitor on uc3842 pin1 contributes to 1 pole, together they provide up to 90 degree phase boost. A large enough phase margin is required to stabilize the power supply.

[@mrfrog8502]

I'm loving these videos. Finally someone explains things properly.

[@JoseSilveira-newhandleforYT]

Probably the best explanation I've seen of the operation of a switch mode power supply. Great work! And thank the cat for pointing to the appropriate part of the schematic with the tail :-)

[@idrisbabatunde6851]

Man! You're a GREAT teacher. Now I understand the basis of single switch forward converter. Thanks my man!

[@rogeronslow1498]

This is a flyback converter not a forward converter.

[@marlonbayaras2850]

Im on a journey of making my own flyback power supply using uc3843, your video really helps, i already made a isolated flyback converter as ground loop isolator, although it only rated for 12-24, now im trying to design the same as in this video

[@ifonlyeverything]

Great explanation. Really helps me understand the UC38xx series, especially the open loop protection.

[@schluderjupp]

Thanks or the in depth explanation. I found it very interesting how this power supply works. I enjoyed this video a lot....and the cat too. 😀

[@kolhatkarchinmay]

When It comes to SMPS this man has got PHD in it! He makes it so interesting to learn power supply and electronics! Have you ever seen cheap chinese supplies where capacitors CAP 's are leaked hole deliberately to save some money on buying large voltage capacitors?

[@daveens5259]

Thanks! This helped me change the output voltage on a similarly designed PC-C-090-130060CT AC Adapter by changing the resister values in the voltage divider.

[@maccheese3006]

I have one of these and I never had the chance to perform some stress testing myself, thank you for sharing!

[@moneydriveshuman]

Still a master piece in 2024...

[@ianuragaggarwal]

Great job in teaching SMPS. Nothing comes close. You have put excellent effort. 🙏

[@pratyushjayachandran]

One of the best explanations on KZbin!

[@bq1000bq]

Beautiful presentation - your help is greatly appreciated. Love from Indonesia

[@Darknight95ful]

some scope images of the output ripple and spikes would be much appreciated in this kind of videos

[@dominikgula1840]

Najlepšie video zo všetkých... Veľmi pekne vysvetlenie princípu a všetkých detailov... Viac takýchto videí prosím :) Inak videá sú všeobecne veľmi dobré a zaujímavé, dobrá práca :)

[@vajkfekete2635]

the clamp has benefits for the heatsink attachment. the bolt can distort the package and have the die separate from the tab internally. some manufacturers tried rivets instead of bolts, even worse.

[@TrentSterling]

Love the kitty co-host! Can't believe the how poorly the heatsink is attached! Really lucky 14 amps didnt blow it up!

[@e1woqf]

Units like this always have to be optimized after purchase. I've got used to open up all electronic stuff I've bought and have a look inside to see what has to be fixed before usage.

[@KuntalGhosh]

@@e1woqf i do this too when i buy cheap stuff , like if a buy a cheap psu i would always check the transistors and board cus sometimes they can put things wrong and cause problem ..

[@Lenny-kt2th]

So a unit that seems to perform well despite some questionable things. To me the datcodes on the Nippon Chemicon main filter cap and the ON Semi control chip seem a bit too old to be genuine items. The unmarked diode and Chengx electrolytics (I thought it was ChongX, the Chinese seem to counterfeit even their own brands) tell the true story why this power supply costs only $8.

[@bruceandmaryvonnecampbell8388]

This is amazing! I am learning fast...I need to because my wife's beautiful, 17 year-old Yamaha Clavinova has stopped working. Just shows a blank green screen, and nothing else, dead as a maggot. So, thank you for the clarity of your presentation. I am in NSW Australia, and I love your accent.

[@e1woqf]

I've got one very similar PSU, probably from the same manufacturer. It works quite nice and reliable for my workbench LED lighting. But there is one really severe drawback which I noticed when I was repairing an old FM Tuner. I was sure that the unit was in good working condition after repair, but I could not tune it to a radio station - until I turned off the LED lighting. So the PSU produces a lot of interference.

[@DolezalPetr]

Mohl bych tě poslouchat celý den, fakt pecka !

[@asthmatic2455]

:))))))

[@zaidhussain5206]

OMG , this is best reverse engineering SMPS details video , this is very informative to me , thank you so much for sharing this video .

[@sorh]

It gave me the shivers not seeing you discharging the main cap :D

[@BenjaminEsposti]

The efficiency of the PSU is actually better than I expected! The capacitors are cheap though ...

[@rogerd4559]

I am really amazed Mr Danyk you havnt built yourself a good lod bank using those resistors instead of hand wiring them together by hand each time. your a modern day Tesler I expect better from you

[@crazycommunistcosmonaut1988]

Great video and surprising how cheap they can manufacture those power supplys

[@martinleska4292]

All components are mounted very straight. Surprisingly :) Looks perfect, but overcurrent protection has more than 40% tollerance. 14Amps it too much and still works. Current sensing resistor needs improvement too. Remove one 0,47R resistor and replace it to 1ohm resistor. I am surprised, because they use 1W resistors or higher in critical places. They used 450V main capacitor. Many cheap power supplies use only 350V caps. They used 2x 75k 1w! start resistors. It could be 150k 0,25W single resistor, but due to high voltage stress this small resistor can break. They designed this power supply for reliability. But absolutely destroy reliability by two screws for power components. Absence of thermal paste is not a problem, because plastic TO220 package can dissipate 30W with proper mounting and heatsing. In this case is probably not more than 5W. Only air gap between package of transistor and diode is problem. Pozdravujem Danyka

[@opera5714]

Little trick. If you reduce the value of the high ohm resistors going to pin 2, this supply will operate at much lower voltages on DC. Of course it will be lower current, but a nice isolated supply for solar. Much safer than a buck converter which are far and few between in the 50-90V range.

[@boonedockjourneyman7979]

Really good work. Great combination of contemporary and vintage circuitry. Amazing depth.

[@DiodeGoneWild]

It's not much of a combination.... it looks entirely so ninetees :).

[@omniyambot9876]

are these type of circuits easy for you?

[@PhaQ2]

The lack of thermal paste is a very good way of insuring failure in the future. With a slight adjustment of the positioning and add thermal transfer paste will likely make this unit last a lifetime.

[@BenjaminEsposti]

It may not need thermal paste, being that it dissipates only 10W. Figure, 0.5W in the diode bridge, 2W in the switching transistor, 5W in the output diode, the rest in various other parts like the transformer and the current sensing resistor.

[@zaprodk]

One thing to note - usually these supplies are built using used parts. For example you can see remains of heatsink paste on the back of the output diode :)

[@barrysteele8132]

Wonderful breakdown. My issue with these power supplies is large components vibrating and breaking their leads. I'll typically secure them somehow. I saw this breakdown as I was modifying the power supply to power a tethered drone; sending the wide-range dc high voltage up the small-gauge tether and regulating it on the drone.

[@Firas95k]

Thanks KZbin for suggesting this channel

[@poptartmcjelly7054]

there should be open loop protection because the IC shuts down when VCC is too high, so in an open loop condition the feedback winding would give too much voltage to the IC supply pin and the PSU would shut down.

[@defaultuser000]

I didn't know you had a KZbin channel. I've built a few SMPSs with the information I learned from your website.

[@nhanle178]

Love your voice, your way to explain the thing, and your teammate:)) I already buy some PSU like that. In my country, it is called "beehives" power supply because of their looks. Thanks for sharing!

[@carloseduardoryffmoreiravi1528]

Great video. Just got this ps here in Brazil with the name Rick 12 120. Exactly same circuitry, and was wondering where to start the safety check before installing it, as it will run 24x7 at 8.5 amps. You saved me time. Thanks and congrats for the great channel and videos. I've been watching it for a while now.

[@krahaborowski]

I just received very similar one, but 48V/4.2A. It has only 50V caps at the exit, while the output can be adjusted higher than that! At least all the components seem to be new.

[@Bob3519]

It is nice to see a fairly well made power supply. (Of course the heat sink issue must be addressed).

[@RoadRunnerMeep]

Amazing work of the schametic and desciptive video, thank you for creating it

[@sirajhussain791]

Nice video on switching Mode Power Supply. Very good and practical way of explaining.

[@bruskqadir6208]

couldn't be explained better. thank you

[@dariofranco3051]

Great explanation on how this smps works. Thanks, Maestro.

[@xddbotnet1370]

Looks like the diode was recycled since it has a screw mark on the tab. But it could have been worse, I bought one of the small laptop ones for a project and not only diode and transistor were recycled and not touching the heatsink, the input capacitor was also recycled and all three were dirty and with remains of heatsink compound they previously had.

[@DiodeGoneWild]

You're right! I didn't even notice the screw mark. So the diode is recycled from e-waste. This also explains why the pins are too short.

[@baharazadi6289]

DIODE GONE WILD-NOTICE please answer my question: is it true in your schematic that in 2 (feedback) connect to the ground directly ? my schematic didn't turn on at all

[@nayib12345]

Excellent video! I'm learning a lot about power supplies from you and finally start to understand the schematics of the complexer ones. Keep it up!

[@annyan904]

very nice explanation.....i wish if you could pls explain each section in detail.....like how the snubber network works the feedback portion....etc.....things will get much clear...also how the two transistor works at the bottom

[@jiouse5881]

Very nice explanation. 👍🏿. Good lesson for unprofessional hobbyist like me👍🏿

[@kpomalproductions5254]

Actually the benefit of smps is that its idle current is very low and doesnt heat when idle. But in transformers for example like 5 amp its idle current will be very high as compared to smps and will heat up!

[@gabiold]

The biggest problem is that it is reactive when not loaded and it is heavier, bigger and because it contains more metal, it is more expensive. And it is not stable, so you have to attach a regulator, further increasing dissipation and complexity. But... When these things does not matter, longevity and reliability is more important, or switching noise can't be tolerated, nothing comes close to a regular transformer. Efficiency also can be higher (usually as size goes up), not particularly hard to reach 90-95%. Fun fact that nowadays when you have to service an equipment, it is 95% that the PSU or some caps are down. When you service an equipment with an 50Hz transformer, it is 99% that neither the PSU nor the caps are at fault.

[@Appregator]

Bless you. Thank you for the reverse engineered schematic. Much appreciated.

[@ilkeozturk7682]

Thanks for your detailed review. I also have the 24V version of this power supply. The 10 ohm resistor on the output diode and the output load resistor (390 ohm) are getting very hot. The temperature reaches almost 80 degrees. Can I use a 100 ohm resistor instead of a 10 ohm resistor and a 3.9k ohm resistor instead of a 390 ohm resistor? Or can you suggest other resistor values to reduce the current flowing through the resistors and thus the temperature? I have the same problem with other Chinese made power supplies I have. But I did not have such a problem with my MeanWell power supply (48V 7.3A), a 5.6k ohm resistor was used as the output load resistor. Is there a formula for calculating the output load resistor? Thank you.

[@omsingharjit]

3 00 how can i modified it for 15 volt ?? I think there should be series resistor with pot maybe by changing that m not sure ?

[@musicasrelaxbrasil]

thank you for dedicating your time in order to share knowledge

[@anibalvilelarj]

Great video ! Congratulations. It is very usefull for manny people. I have used it to modify the range of voltage adjust of my power supply. Thank you.

[@robotics_and_stuff]

Recently found your channel. I'm very satisfied!

[@ergindemir7366]

What kind of range would you get if you swap adjustment resistors, and output capacitor? What is the minimal voltage you would get?

[@luich1288]

The primary capacitor is a nice nippon chemi-con (quite expensive for this price, i'm surprised)

[@sharatsaxenaindia]

It was a learning experience to watch your video. Thank you very much.

[@lastawake2822]

Maan, you're pretty great. Thank you so much for explaining it so clearly. It helps me A LOT while i'm learning that kind of stuff. High five, my close neighbour, high five comes to you from Poland! ;)

[@laptopdisassembly9886]

This is exactly what I was looking for, thanks !!

[@brahimabdelsadek9534]

Genius you are man.thanx million time big respect from algeria.you are the best teacher.

[@465kuba]

Great video. I'd like to see a movie with an oscilloscope, where you testing power supplies.

[@tcirit]

I would like to see your own project "High power adjustable switching power supply (SMPS) 3-60V 40A" with 10-20A version of the PCB files and parts list. Thanks a lot for this video and your great effort!

[@PerrynBecky]

Beautiful cat! I love the markings.

[@xcsourceindustrialist7926]

holyshit... you are the owner danyk.cz ?? i always thought the owner of danyk.cz is someone who are 60-70Yrs old in full of experience. that website even on my bookmark in every pc i had. you are tie for fully expert in circuit explanation to Elliot Sound Project. that 2 websites is very determining my memory of electronic. as newbie for making electronic project you are like GOD... lol, cheers mate

[@lvb1b463]

A great power supply for not only LEDs but also Halogen and Incandescent lamps.I have a question though. Can it supply power to a HID (High Intensity Discharge) lamps?

[@MicheIIePucca]

I had purchased that same power supply so this really interested me! Great video... thank you!

[@kagabap]

i like your accent because i can easily understand your explanation,thanks.

[@Fifury161]

Great work, fantastic presentation and formulas - thanks for sharing!

[@dama4357]

even more resistor is the new even more salt? :P

[@DiodeGoneWild]

It may be :)

[@kimt6333]

It's not an industrial SMPS, just open frame, and they are not limited to industrial installations. Besides that you will pay way more for an actual industrial power-supply of any sort.

[@SriniVaith]

Wonderful explanation of smps.Thanks.after the switching transformer how is pulse output converted to dc?

[@amw224]

Very clever circuit for protection if opto is out. Is it work as current protection too?

[@honda9731]

very nice explanation DiodeGoneWild. Hope you also do Cordless Tools Battery chargers voltage conversion from 120V to 240V such as Ryobi, Ridgid (AEG), Kobalt. These chargers don't have voltage dividers (after the bridge rectifier), so I don't know if conversion is possible.

[@DiodeGoneWild]

I have never seen those :). I am surprised that purely 120V switching power supplies are still being made those days :).

[@ELECTROHAXZ]

You always make the best videos. Great job and cute cat 🐈 :)

[@technology520]

Hello! DiodeGoneWild.I have my 480W-12V smps. The name on the 8 pin pwm controller on the circuit board cannot be read.I looked through a magnifying glass, but I couldn't see it.How do I know what this pwm controller is?I have a digital multimeter. Not much electronic knowledge.Please help me with this.

[@chimpandolfo2008]

Muchas gracias por tan explendida forma de enseñarnos tanto en prueba como teóricamente el funcionamiento de la fuente!!! bravo!!!!

[@nashkmr]

Can someone please clarify? The transformer seems to be on the DC side. I've learnt transformers work on AC only. What am I missing?

[@fly4fun24]

i have a question about my 30v 2A recliner power supply. As soon as i pull more than 400 ma the voltage drops to 0 and comes back after I remove the load.. i replaced the optical and the TL431. No luck

[@r3nx0]

hello excellent video, but I have a doubt, at minute 3:49 you measure the input power using a multimeter but I see that you use the mV range, I would like to know how to measure watts with a simple multimeter, thanks

[@sri198904]

Nice video !! Are you sure the efficiency of flyback without synchronous switching can reach 92 % ? even the best buck and boost converters in the marker struggle to reach 90 % without synchronous switching and here we have an isolated flyback topology.