i have been using it and it’s been fine
but it buzzes now
and also cuts out if you draw more than low power
note
THIS IS NOT THE POWER SUPPLY I FIXED BEFORE it’s different
anyway it’s rated for 10A but is dying being asked for intermittent 6.5
and nothing is obviously hot
so yeah
what’s it likely to be? power transistors?
@moira The buzzing is probably coil whine in the transformer, from some subharmonic of the switching frequency that happens to be in the audible range.
It's possible there's a bad capacitor somewhere, like in the feedback for the controller, or if the power transistors have a snubber circuit.
@toroidalcore it seemed too low pitch for transformer whine but I didn't think about lower harmonics so maybe that's it
capacitors are easy to check, at least. ^_^
(this is not the same model as the one with the splody cap and the values are all different so ... sure hope I have some lol)
thanks ^_^
@moira You can get weird effects if something goes a little bit unstable, or unstable during certain load conditions. It's kind of hard to say without poking around with a meter or scope.
Good luck!
@toroidalcore @moira if it sounds more like buzzing, it could be the core laminations becoming, er, de-laminated. try squeezing the core (in the direction that squeezes the laminations together) and see if that helps at all.
admittedly this is more common with older, larger transformers.
@wohali @toroidalcore oh it's buzzy for sure
i'm kind of afraid of operating it with the case off given the previous [POW] incident tho'
@toroidalcore @wohali well guess what's not making the noise and not cutting out anymore
so it must be heat dependant and there must've been some heat even tho' i didn't feel any
@toroidalcore @wohali WHELP we're back
big voltage drop and the buzz, gimmie a sec for recording
@toroidalcore @wohali and it's leaking 60hz (at very low AC voltage) at all times, I didn't check that before
also I thought the buzz was higher pitched than I'm hearing now but I could be just remembering that wrong
@toroidalcore @wohali It's the same either way, so this part is a consistent problem. I ... hm, can I zoom in on that? _Sorta_. Lemmie try things.
(it's a very silly little oscilloscope. I should get a better one.)
@moira @toroidalcore that's not terrible AC overlay, but it could be improved with better output filtering.
i assume you're not getting the voltage you expect here? or is it just a buzzing?
as always with these things, suspect the electrolytic capacitors. look for any leakage or funny smell.
here's a very good intro video to how SMPS work if you need one: https://www.youtube.com/watch?v=3FyXqNcqvRM
@wohali @toroidalcore oh now when it buzzes it's dropping voltage like a mofo
and quick, too
this is all new, none of this was happening until this afternoon so something yearns to fuck off into the sea
@wohali @toroidalcore oo, someone wants me _alive_ this time, that _is_ new
xD
the tiny scope is hilarious and only battery so we are 1000% safe there
this hasn't been running not, it's been running quite cool which is why the big surprise. it also only has a couple of weeks of use on it, tho' many hours a day during those weeks.
the 6.5A draw is barely more than momentary, most of the time it's drawing .21A according to my bench supply.
The big cap says "nichicon" but I'm rolling to disbelieve. the rest seem to be nonames
@moira @toroidalcore there's not a lot going on in this design.
always suspect the electrolytic caps first. especially if this has been running hot somewhere, they'll be the first component likely to fail - and are often the first component cheaped out on by inexpensive contract manufacturers (China and elsewhere).
if you have a cap tester i'd pull each of the electrolytic caps and run them through it. be sure to discharge them first! you can probably skip C1 for now unless it smells or shows obvious signs of failure.
@wohali @toroidalcore the two little caps near single IC (very similar, separated by resistors) are showing weirdly high ESR in-circuit so that's fishy
the "nichicon" C1 is showing less than half its rated capacity in circuit which is also kinda sus
(I have a dedicated ESR meter, PEAK Atlas ESR70 Plus, sometimes you can get hints from in-circuit components with it)
@moira @toroidalcore ah, ok, i don't expect everyone to have an ESR meter, esp. not an in-circuit one.
the caps near the meter there are may not be as critical, esp. if they are just decoupling for Vcc on pin 7. (the 8-pin package won't have pins 11 and 12 broken out, they'll be connected internally.)
C1 is going to be the post-rectifier filter on the input side and, while it'd be interesting to look at, is probably something you're not setup to probe with your current gear?
you could check that the IC is at least getting the power it expects at pin 7 safely, if that's all over the place the controller isn't going to be doing the right thing at all.
@wohali @toroidalcore the esr is one of the relatively few fancy tools I have
both it and my true RMS multimeter say this 150uF cap is more like 70uF so that seems kinda bad
@moira @toroidalcore i do my best not to assume anything! but i also do my best not to talk down. it's clear you are not clueless!
i only mention the leg-lifting because it's my standard practice. with in-circuit ESR you need to have some confidence you know the inductance of the rest of the circuit from the capacitor's POV, that it is not paralleld with anything else that will induce a reluctance, that the 20mA or so it injects for larger caps isn't going to do weird things, and understand that e.g. in the ESR70+'s case it measures at 100kHz, which is not at all the frequency value that matters for C1 (that'd be 60Hz). i believe the newer ESR70 "Gold" measures at 50kHz, and devices like the DE5000 (and other LCR meters) have selectable frequencies.
hope the brain dump was enjoyable, now back to cookies 🍪
@wohali @toroidalcore I like you already so I'm not mad it's okay xD
There are two others of these sitting around old-stock-in-box (all 2020, yes these were get-what-you-can-get plague purchases lol) so I grabbed another to make comparisons and in circuit and the equivalent cap on the working one actually measures as labelled even still on the board. So this one is clearly a nonsense boi
tho' this does beg the question of "how did this work for a while?"
still want to fix it tho', if I can
@toroidalcore @wohali Possibly - they're still testing basically the same across both units tho'.
I also checked all the diodes while I was in there and they're testing fine in-circuit except for D8 which I can't actually test because it's in parallel with a resistor so probably a noise-reducer I think? But it looks okay for whatever that's worth
@wohali @toroidalcore I have not yet watched the video so I do not yet know what that is xD
one of the power transistor's just off to the left if that means anything
@wohali @toroidalcore also this "nichicon" weights absolutely nothing
i wonder if there's literally a smaller capacitor inside this bigger one xD
Solarbird 
ToroidalCore
wohali