I repaired a Power Supply using cold air

[@rickchagouri-brindle6520]

New technique to me, thanks for sharing. Never too old to learn new tricks.

[@tony359]

You're welcome!

[@fast_squirrel]

Great video, Tony, and great job fixing the PSU! Wasn't aware of the diagnostics method with the compressed air, very nice to know.

[@tony359]

Thank you! Very often you have boards which only work when cold or when hot - cold spray and hot air might help in triggering those issues. It's a pretty important tool to have! Thanks for watching!

[@ted-b]

Nice work Tony, methodical.

[@tony359]

Thanks Ted!

[@FireballXL55]

Hi Tony I like your down to earth video's. I do quite a few medical grade PSU's mostly from ophthalmic equipment, you could use IPA as it would evaporate and cool the device you are just not looking for the quick evaporation, I just use freezer spray as I have never owned any canned air, I have a compressor to blow off all the dust. I guess I have been lucky and not had a nasty one like that, they usually won't start or are blown. I also use a 15W lamp to discharge the main cap, much better than the palm of your hand when turning it over. 😊

[@tony359]

Hello and thank you for your comment! I also use a 15W light bulb for ovens to discharge caps! I always keep some canned air for delicate jobs or where the dirty air from my compressor is not suitable. Plus, it can be used as freezing gas 🙂 thanks for watching!

[@samuraidriver4x4]

I have used a bit of heat to pin point intermittent issues before but never thought about the other way around. Heating up specific components from a cold state also gives a ball park where to look but cooling it might be a better way. And again it was a tantalum capacitor again, people always complain about electrolytic caps but the tantalums are worse if you ask me.

[@tony359]

Hot air is a good idea indeed. I did not show it on video but to "accelerate" the issue I have also warmed up the board with hot air. Both hot air and cold spray are valuable tools when it comes to electronics! Thanks for watching!

[@Sydney268]

Smart work, well done - was I the first to notice Mulder and Scully? 😂

[@tony359]

You are 🙂 Thanks for watching!

[@sudheerm.s6792]

Thanks for the elaborated explanation 🎉

[@tony359]

Thank you for watching!

[@Epictronics1]

The mystery machine fixed! :)

[@tony359]

Very mysterious indeed! :)

[@yvesrochet3288]

Tony, beautiful operation, i think you are using butane, same as lighter for smoking, be careful which spark. Thks

[@tony359]

Yes it is the flammable one - it was cheaper :) Absolutely, might not be a good idea on live PCBs with mains, thanks for that!

[@ziko2371]

Nice video, cold spray, is it as I know freeze spray -40° ? I use freeze spray sometimes when programmer not connect to MCU.

[@tony359]

I use canned air, upside down 🙂 Not sure what temperature it goes to be honest! Thanks for watching!

[@vswitchzero]

Great video, Tony! I've never tried this method but I've heard people suggest using it to identify bad memory chips on graphics cards too. Apparently you spray individual chips until the on-screen artifacting changes and you've found your bad chip. I will have to give this a try some day! Any concerns with the conductivity of the spray? or does it evaporate very quickly and its not a problem? Cheers! 👍

[@tony359]

That is indeed a primary tool in my toolbox and can be used on a number of issues 🙂 The Gas is not conductive. The layer of water which forms on the surfaces should be pure H20. Pure water is not conductive - the salts dissolved in it are. So condensation should be safe, but you always have the risk of contamination from the PCB itself. I don’t see it as a problem though. One thing I should have done is to use a non-flammable spray can. Thanks for watching!

[@vswitchzero]

@@tony359 That makes sense, thanks Tony! I will definitely give it a try one of these days 👍

[@baghdadiabdellatif1581]

Nice. Did you see any thing on that cap under thermal camera? Did you use hot air on that cap to make that intermittent problem appear faster?

[@tony359]

Well, if you watch the first time I used the thermal camera, you can see the cap getting hot - but not as hot as the diode which stole the scene! :) The PSU was protecting itself so the capacitor did not explode or get too hot. But the diode, being a smaller component, did and that fooled me! :) No, I did not use hot air on this occasion. Thanks for watching!

[@christopherjackson2157]

I've seen ppl use resistors to drain mains caps on psu's just to be sure. Not sure what spec is required for this use case, but I've meaning to look it up - Must be better than my current method of just leaving the psu disconnected for a week before opening it lol Definitely always check the cap with a multimeter before working on it just to be safe 😊

[@tony359]

I use an oven light bulb. I soldered a couple of wires to it and insulated with tape. It acts like a powerful resistor, which increases resistance when it warms up. Ideally you don't want to short the capacitors but you don't want to wait 5 minutes with a resistor in your hands. And the resistor shall not burn. I like the light bulb option as it also gives you a visual confirmation of the charge being discharged. Then, as you say, you can double check with a multimeter. Thanks for watching!

[@christopherjackson2157]

@@tony359 good idea. I'll give that a try :)

[@sokoloft3]

I think you can do the same with isopropyl. It will evaporate faster in heated areas.

[@tony359]

That's a good tool but it's for something else. With isopropyl you can identify a shorted component which is burning hot. That capacitor was not shorted, it just barely worked but only when cold. As soon as it would warm up a bit, it would stop working. Cooling it down with liquid gas would allow me to "make it work" so I could identify it. I hope this makes sense! Thanks for watching!

[@sokoloft3]

@@tony359 Ah okay I understand now. I had seen a curiousmarc video where they used this can air method to find a broken via. Thank you for the explanation and video!

[@tony359]

I watched that video! The one where they use the magic finger :)

[@SanelKeys]

I guess +/-15 V comes from 7815 and 7915, so only when input voltage becomes less than 3V higher (in absolute value) than output, i.e duty cycle is low enough, 15 V starts to drop. By the way, I like a lot this kind of repairs, SMPS. Keep up with good work.

[@tony359]

+5V and +15V are definitely regulated by the main controller with feedback via optocouplers. Maybe -15V has a linear regulator, I did not check. If there is one, indeed it works as you say. Thank you for watching, I'm glad you enjoy this kind of content!

[@menotyou8369]

Highly unlikely that voltage rails from a 7815 or a 7915 would have adjustments.

[@sudheerm.s6792]

6:00 multiverse reference 😊

[@tony359]

I think I had watched the movie just days before recording the video 🙂

[@sudheerm.s6792]

@@tony359 yeah. Me too. Fan of movies and influenced by movies

[@sudheerm.s6792]

❤

[@CooLDEaFY2023]

Cool cool cool

[@tony359]

Speaking of freezing gas :)

[@Bergi2000]

Nice done! 😊

[@larrybud]

15:20 the "111" is a line pointing to which terminal is positive.

[@tony359]

I would have never guessed that! Thank you!

[@tassdesu]

By the way, what will happen if an arc or spark occurs at the time of using the spray with the power supply turned on?

[@tony359]

Probably nothing good! :) You can buy non-flammable sprays and I should have probably used one of those myself!

[@SilliestSmurfEver]

3:44 pointing directly at a cold solder joint, 15:51 it´s gone😁

[@tony359]

I think it's just cleaned, I don't think it was a cold joint, just a factory joint which hadn't been cleaned as it often happens. Thanks for watching!

[@SilliestSmurfEver]

​ @tony359 Thank you for your reply. The middle of the three solder joints next to each other horizontally looks like it has a ring around the leg of the component, to me that looks like a broken cold joint. Thank you and have a nice day👋

[@tony359]

I see what you mean, I often see non-existent cracks when inspecting from above. It might be though, I don't know, it's been a while! But thanks for mentioning! :)

[@SilliestSmurfEver]

@tony359 Yes, that's true. It may or may not be. Thanks for your kind answer and have fun with your projects.🙂

[@simontay4851]

Surprise surprise, its a bloody tantalum again. Little buggers are so unreliable. Worse than 'lytics even. You should also check the other tantalum caps. Compare the ESR of a new 1uF electrolytic to this. 6 ohms is quite a high ESR.

[@tony359]

You know what, I was wondering that indeed. Now, that little tester is not super reliable when it comes to ESR but it caught my attention. That was the brand new one BTW 🙂

[@8bitbubsy]

Tantalums are not considered unreliable. It's easy to get that impression when you watch too many repair videos on KZbin. It's especially common that they short when they are very old, but that's WAY past their expected lifetime anyway. Like all types of components, they can randomly fail.

[@catriona_drummond]

suppose you need to be wary of condensation though, when doing this. Don't want to create shorts.

[@tony359]

As far as I know, water is non conductive. It's the salts in it which can make it conductive. Condensation should be pure H2O so non conductive. Granted, the condensation might pick up whatever contaminant is on the board. But I don't think it's a massive issue here to be fair. Maybe with higher voltages? Thanks for the feedback though!

[@clintcolombin]

I don't blame you for not showing the device. Medical device companies will litigate to prevent 3rd party repair.

[@tony359]

Thanks!