12V 50A 600W power supply - failure analysis & transformer autopsy
ฝัง
- เผยแพร่เมื่อ 7 ก.พ. 2021
- Today, let's analyze what failed in the cheap dodgy 12V 50A 600W switching power supply module from eBay from China. It stopped working after 1 hour and 7 minutes at full 50A load. Let's find out what components are faulty and let's see what's inside the switching transformer.
Next episode - rewinding the transformer
• 12V 50A 600W power sup...
Previous episode - the test of this power supply:
• 12V 50A 600W power sup...
First episode - the teardown of this power supply:
• 12V 50A 600W power sup...
Please support me on Patreon:
/ diodegonewild
My Instagram:
/ savage_danyk - วิทยาศาสตร์และเทคโนโลยี
![[TH] VCT Masters Shanghai Playoffs DAY 2 // PRX vs G2 | EDG vs TH](http://i.ytimg.com/vi/T5j2XVcDFs0/mqdefault.jpg)
This is massively useful content :) I don't think there is anywhere else you will see such a well done practical analysis of smps's. I've worked with many engineers who could write a board full of formulas about power supplies but couldn't repair one if their life depended on it. This sort of practical knowledge is rare and getting rarer. A massive thank you to Diode for these videos.
Absolutely true
I couldn't say it any better! Thumbs up for DGW!
Spot on ..........
Absolutely agree!
Absolutely 💯 right 👍. My well wishes to Gentleman 💗
Wind a new transformer to show how a good transformer should be winded.
yes
yes please
Yaaap
Yes
PLEASE
1:20 They must have used a very high quality fuse that could withstand a 25A current! 🤣
Thank you for yet another very interesting and excellent video!
Thank you. Would be nice to inspect the fuse and see how many amps it takes to blow it up. Its probably a fake fuse too.
I think, they used some pice of the primary winding - solid aluminium is best for 25A ! :O)
That's possible - i've encountered similar situations in high-end power supplies. These fuses are just slower than good circuit breaker.
@@michamaecki8104
The 600W power supply tripped a 25A 230V circuit breaker (5750W) and it took a few seconds. 🤔
@@Conservator. Yes, because fault initially didn't blew up transistor, that shorted mains. First transformer failed, and after 5s transistor and diode. Also, circuit breakers have solenoid that trips breaker at huge current surge much faster than ordinary fuse, so it's common to see power supplies with blown transistors and fuse intact.
Also, to trip standard "B" circuit breaker in 5s, ti would need to see something around 75A.
To rule out secondary side shortcircuits it's useful to plug a bench power supply at the outputs and see if there's any abnormal current draw. This can also be done at the input but a higher voltage is needed (above 40-60V DC).
BTW these SMPS videos are absolute goldmine, I have the same ESR meter but it never occured to me that I can test transformers with it. I'm now trying to build a TSWF power supply.
thats a great tip you shared there. thank you Bogdan : )
Would love to see a repair of the power supply with explanation how to choose the right (real copper this time) wire for the transformer.
Would love to see that back in operation with a hand wound transformer with copper wire used. Bit of a pain I known, but would be nice.
+
i have being dreaming with that day since i born
Hell yeah!
He did
@@krisd1000 Yes, I watched the video moments after he released it.
DiodeGoneWild's birthday is in three days!
😃😃and my on 18th...
Tak že, dopredu všetko najlepšie majstre.
😀
Mine in may 5 2007.
How do you know 🤔
@@Purple431 mine may 13 2005
i would love to see this working with rewinded chopper transformer.
Will you fix it, I mean make a new transformer?
+
no fix :/ I would be verry happy to see you rewind it with old copper wires :D, as always, good video dgw!
+
It seems like the usual curse of cheap Chinese products: passable design, horrible execution.
Had they used a transformer with copper windings and slightly better input caps (better meaning not smashed to smithereens in this case), odds are that the PSU could have worked for years.
or given it a realistic rating
Passable design probably copied from other vendor's product.
Which is the ongoing problem with Chinese products
@@jms019 I'm sure this was a 25-30A power supply they stuck a "50A" label on
Informative video. He may be the most resourceful technical person I've seen. And the dim bulb tester is a great tool to observe the power cycling.
I was about to ask if you have any videos about air gaps in transformers, but then i reached the end and got an excellent explanation!
Great video!
Now that's a nice SMPS video. I didn't think they would cheat even on the copper wire! I was wrong again.
From the output rating to the parts used, all the cheapest possible.
I am always impressed by your deep knowledge of the principles of fundamental electronics.
Great work 👍👍👍
Great analysis DGW ! Those impedance tests and demos really are educational :o)
This sort of vid should ABSOLUTELY be shown in electronics schools. Good job mate. As an electronics numpty I love your vids anyway, but I can really see the educational value here.
Please wind the transformer.. the supply needs it
+
Great video!
If over the next few episodes you can walk us through transformer design, winding, testing, that would be extremely appreciated and valuable.
Upvoted
Thanks my friend for sacrificing your power supply and explaining the fault.❤
When you click the "like" button on a video before even seeing it, it's when you hit gold...
One thing you said has me thinking. That a flyback design has an air-gapped transformer. My understanding is that it allows more energy to be stored for a given size (amount of core iron) and the nature of flybacks requires more energy storage. This is what I like about this chammel. Despite decades of experience there's always something new to learn
I'm learning a lot thanks to your tutorials
other comments: please build another transformer for this psu
me: please build another transformer for this psu
diodegonewild: schematic first
+
Very educational video. You've made a huge effort. Thank you
Very very valuable information. Thank you.
Thank you dear for your great SMPS videos, I learned a lot of information about them... I can’t wait the reverse engineered schematic of it
Since there were no shortage in the primary side, I would have tried the power supply without any current-limiting bulb, just in case the SMPS controller was disturbed by the voltage drop across the bulb. But you were 100% right, the transformer was faulty.
Also, nice trick to test the bridge rectifier directly through the power cord.
Very good explanation video
The SMPS controller only need a small voltage and current and DGW suspected the primary side of the transformer to be shorted and didn't want to fry again the chopper transistor and the diode again and therefore used a current limitation bulb
@@emmanuelvallee5386 lol tu es là toi !! excellent
Very nice analysis and explanations, thank you very much !
Really wanted to see you rewind the transformer properly and then re-test it at 50A!
DiodeGoneWild Happy birthday!
Thanks :) How did you know?
@@DiodeGoneWild
😁 Someone commented that three days ago.
I hope you’ll have a wondered day!
(That someone was Václav Trpišovsky)
Excellent review as always.
Highly informative and important to us this knowledge about smps transformers.
Very genius!!! Amazing the things I learn here. And I like your cat, of course.
now a tutorial on its driver chip please. this is top 3 best electronics tutorial channels
Love your accent... and content is absolutely top notch!!!
Very good knowledge of electronics
Very helpful video 👍
Very interesting and informative presentation. Lovely cat.
Thanks Dany very interesting
well explained!! a thermal camera (flir) really helps on these troubleshootings but it is very expensive..
i interested on this topology since i build my two switch forward welding inverter based on your schematic. Cant wait for you next schematic video
Great series!
Great information thank you.
I'm sure I'm not the first one to say it, but you have an absolutely captivating accent ;) Thought you an Indian at first. Awesome videos, keep it on!
Aluminium windings, what a con :-(
I didn't expect that.
Your cat would make a better supply :-D
CCA wire is now so common in Chinese products that getting something without it is very rare now. They make it thinner and thinner, I suspect we will soon see all wire coming out from there in products being CCA wire soon, from power cables to the magnetics, with the only copper being PCB traces. The plugs and sockets they supply are getting thinner and thinner, with the metal content being reduced every year, and soon they will probably be going to copper or nickel plated steel for contacts in all products.
@@SeanBZA I had noticed mains power cables using coated aluminium but if the use is light at 240 volts it won't make a massive difference.
It looks like we will all have to hoard real mains cables, ok i do anyway lol.
I bet you have a few hanging around from previous computers and such.
@@SeanBZA Tehnics uses CCAW on speakers for at least 20yrs
Nice Video! Good explanation :)
DGW is a clever, knowledgeable man !
So enjoyable to watch 😊
The explaining i look forward for! That forward switching is not totally clear to me. And you can abnormally good and clear explain such things. (Like that second outputdiode that works together whit that outputinductor, you made it totally clear to understand in just one sentence!)
HAPPY BIRTHDAY!!!!
I'd also like to see you repair it with a rewound transformer. It's a great opportunity to explain how to design one for a specific job. I hope the schematic explains what the missing components marked on the board are.
I like your makeshift thermometer
:)
EXCELLENT sir !
Was intresting to see you using a esr meter to check the transformer.. what do you lookfor when usin this method...i have this same meter
Thanks for this vedio.
why the impedance is high with a frequency of 100khz delivered by your esr meter
Is it possible that the insulating coating can function to allow the skin effect to take place? The tape will function to dissipate stray inductances? Thank you for the information on improved coupling. I have found resistive wire to handle heat. Perhaps the resistance/heat diminishes an inductive essence. What material was the coil inner written? If it is iron based ferrite, I would appreciate it's conductance as improving voltage capacity, and coupling for the coiled inductance as well.
Great video! Aluminium wire is the devil's work. Microwave oven transformers are wound with aluminium as well, but they don't even use copper plated wire. They use copper coloured varnish to make it look like copper wire.
Imteresting video. I’m surprised nothing more than the mosfet and diode failed after the transformer shorted. It would be good to see if you can re-wimd the transformer with properr copper wire and see how much better the performance is, you just need to find a new bobbin. After that you’ll have a useful worrking power supply to use for something else.
Great thank you fo these videos!!!
👍👍 thanks for the info
thank you very much for sharing
I really love you your videos sir you great and your explanation
More power supply videos please I love them 😘😘🤩
wow, that was great I never would have suspected aluminum wires in the transformer... I have a old 60A bench top power supply from 1950's that has a huge 10 inch transformer in it that uses aluminum wires but that was due to copper shortage because of the war. if I leave it on for long times it will get warm with no load...
interesting! maybe some capacitor or diode has been expire thats why its get hotter
Thanks so much @Diode gone wild. Please share schematics.
After watching a lot of your videos encountering doggy components and constructions, I decided to open my 48V 10A power supply and it turns out to be a BJT half bridge. However, my power supply is welly assembled, with no doggy components, and the semiconductors are mounted properly with heatsink compound. Unfortunately, I found out that the snubber resistors next to the secondary side diode are getting burning hot, so I replaced them with higher power ones and replaced output smoothing capacitors with low ESR ones. The only thing I found doggy was that they completely bypassed the inrush NTC thermistor...
please build a transformer for this power supply in order to make it work again! thanks
+
Curious if it can achieve the rated output with good quality high temp enameled copper instead of aluminum. Though not sure how much better it would be due to skin effect. Probably a good idea to use thin wire, like 35AWG and like 5 strands of it to make the best use of the copper.
Hello, i have a uc3843 with very low switching frequency and then only have 1.65v at the exit. Any idea where to look at ?
hello, can please make a dedicated video on smps transformer designing?
because you explain it very well
and i liked your video
Happy Birthday Diode or D_____L 11-2-2021 you success fully survived 34 year..
Yes and thanks to @Václav Trpišovský for telling he's birth date.😅😃
U should rewind the transformer and fix the dojy cheap chinese power supply...this will be interesting😊
I was curious if we can convert a flyback smps supply from 0-12 volt.. The minimum i got was 3.3 volt but i need 0-12 volt... Is it possible i can give external supply to main switching control ic ... Will tl431 and other circuit work with it ??? Share your views.... It would highly appreciated... Thanks.
pls keep uploading videos like this
My desktop PSU failed and I found a blown chip inside TNY280PN, what does this do and should it be possible to just replace the chip?
@DiodeGoneWild It is normal, wires may be stucked together. Why? Because SMPS transformers are winded and covered with insulation lackquer, just like classic, 50Hz, one... I disassembled good EI-33 transformer, and windings were also together and I noticed lacquer residue...
Hello sir , active speaker is rated 1500 watts , at it seems the primary side is blown already cause of over voltage , the driver is ir2153 and no opto design , is there any way to mod the circuit?? Plese help men
and just because of an aluminum 'copper' wire, that the design failed. too bad. thats why im a bit skeptical about these cheap high power switchmodes
I wouldn't say it would have been fine with the same diameter copper wire. The copper coating actually helps more than you may think because in a HF transformer the skin effect causes most of the current to run in the outer layer anyway, so not as much current wants to flow inside the aluminium anyway.
So it depends on the thickness of the copper coating and of course the frequency how much better pure copper would be.
hi Mr. Wild, I build two switches forward converter, with feedback to regulate at 10 V output. When I did load test, it was good till 1 amps. Between 1-1.8 amps it still good with annoying noise. But when it came to near 2 amps, both mosfets fail, I was prety pissed off to loose good mosfets that unfortunately costs extremely expensive in my country. What could go wrong? does it has something to do with high dv/dt ?
sorry for my bad english
Greetings from Philippines
You should really get one of those Atmega328-based transistor testers. They are super cheap and can measure a lot more than your capacitor tester.
DiodeGoneWild what would cause a computer 19.5 v adapter to be showing 19.5 v when test but no able to power the computer ?
very interesting. QUESTION: Will you add a proper transformer and heat sinking for the secondary diodes, & add components ETC to make it a proper 600 WATT PSU??? there isn't much work need to do it! and id bet you have the parts laying around you would need for it. so will you consider it???
Love it. 🙃👍
I'm amazed by your deep knowledge and teaching skills.
Do you have a method to measure parameters of transformer/inductor cores? During the years I have collected a lot of old cores but after some time you never know where those came from and what type they are - ferrite, powder or whatever else, to be able to re-use them in some way. I am just not able to throw them away, i think I am not the only one.
I have seen some methods in russian welder builders' sites somewhere but can not find that any more.
EE or EI cores are probably always ferrite. The rings are tricky, they are often iron powder, but ferrite too. But they tend to have color codes. For example the output inductors of ATX tend to be yellow with one side white or green with one side blue. There are tables for those color markings. And of course you can wind a certain number of turns on it and measure the inductance. This gives you a rough idea about the permeability. You can also measure the effective cross section of the core. Mostly from the diameter of the centre column or from the two dimensions if the cross section is rectangular. And of course the air gap is visible, this tells you it was a flyback or maybe a PFC inductor. A ferrite core with no air gap means a forward or (half)bridge. But I tend to mark transformers I take from something. For example I mark it "100W 12V 80kHz flyback, no PFC". If I keep just the core with no winding, I mark it lets say "100W 80kHz flyback, primary was 90 turns". This makes it easy to reuse the core for my power supply, without having to do much calculations and without trying to guess the permeability and the maximum magnetic induction which doesn't yet saturate it.
I bet the diode gone wild
First I downloaded the video and watched it. Now I came back to comment you should rewind the transformer.
Only a few numbers turn so it will be much easier.
Greetings from Philippines..
Idol
@@JPDicamVlogs andito ka din pla hehe
Hello. Thanks for all your great Videos!! A little question, or rather an idea: Maybe you could explain a escotter controller, i.e. its construction and functionality in detail. For example, one of the models that are available from Aliexpress for the large escooters, i.e. those with 25-45Amps, would be good. I think many would like to understand how it works. At the moment e-mobility is very topical and many buy something like me, but unfortunately understand very little about it. god bless you and keep it up !! Thank you!
Subbed since 2016
Hi !
I was so scared when you touched the live part using your finger. It s not a voltmeter or a Thermometer. ;)
Do you use an insulation transformer ?
Look the power cables he uses;
It is French standard that is grounded and polarised. He was breaking the live wire, so only ground and neutral staid connected for sure, if the all sockets were wired correctly.
In that case, no danger.
Weird how the primary OCP didn't prevent the obliteration of the diode and transistor, but works fine once you replaced the blown components and just the transformer primary shorted.
I think that the transistors and diodes would fail again if I left it on a bit longer... The overcurrent protection has a certain delay and with a very little impedance in the primary, the current rises very steeply when the transistors turn on. And the diodes have no heatsink, so they tend to overheat before the transistors do.
@@DiodeGoneWild Typically the OCP circuit has a RC circuit on the input, to handle switching on transients that might cause premature switch off, so the current at turn on can rise fairly high, probably well past the SOA ratings of the transistors and diodes, with the shorted transformer. The overcurrent protection is there to normally handlke shorted output, where it will be transformed to a rising current in the primary, and also as cycle termination in normal operation, so it only has to be really operating near the end of the conduction cycle.
Shorted primary the current rises very fast, and the turn off only occurs after the RC delay, so current is going to be limited only by ESR of the capacitors and the on state resistance of the MOSFETS, the dodgy capacitors here might ironically have enabled it to survive, as ERS would have been in the tens of ohms, limiting current pulses, though the output voltage would certainly not have been 12V at full load, dropping to whatever the capacitors allowed, with a very high mains ripple component as well. Diodes might not have failed, but the capacitors would definitely have gone boom and given a confetti shower.
I feel like I've seen power supplies that looked like this and were rated something like 30A for the 12V version. Did this unit have over current protection?
there is fuse in it but its not design to detect overheated, probably the transformer dies slowly by overheating and finally failed to give output. the interesting is there is no component blow up or burn at more 100 degree
@@xicath I know that. I'm talking about how most of these Meanwell type supplies, being idiot proof industrial electronics, will reset if shorted or loaded beyond rating, as it was after the repair attempt. So did the extra shady manufacturer sub a copper transformer with high temp wire for crappy CCA, or just alter the current monitoring circuitry of a lower power design? On further thought I'm guessing the former, or both.
@@Broken_Yugo there is meanwell made in taiwan, i think its have better components than meanwell made in chinese
How can you tell which side of transformer is primary or secondary??
Which topology is this ? is that forward converter ?
Would be nice to make a video with this restored transformer. Of couse with real copper windings. ;)
+
and the lesson taken from this for china power supplys?? pick a higher power rating than you need and not push to max or rewind the transformer with copper wire and perhaps the Inductor. Rewind with copper for these supplys might be a good video we all can benefit from
I know that its impractical, but secretly hoped that the dodgy components could be repaired and meet the specs safely. Maybe in a wind tunnel :) And I don't think I've ever said dodgy before watching this series.
Great Job my man! Please can you show schematic?