Reducing/removing hissing noise from old PSU (1982)

[AndreyS]

I've got old a computer: Highscreen Colani 486DX-50. And it starts hissing when PC it on for 2-4 minutes, that is hissing noise slowly increases over time.
Pitch and character of the noise is changing when under load.
The PC is made in 1993, but the label on the PSU says manufacturing date is 82.02.19. I cannot get how it is possible.
1) I desoldered some capacitors, which were made in Japan (Jamicon) and they are still good, replaced with new low ESR caps, which have worse ESR. PSU still hisses.
2) replaced ceramic caps with another - hiss remains
3) tool audiorecorder, headphones and lavalier mic, then searched for the hissing source, and found that it is most likely the bigest transformer

My questions are:
1) should I replece those caps?
2) how to reduce/elimenate hissing noise from the transformer?

[wraper]

1) I desoldered some capacitors, which were made in Japan (Jamicon) and they are still good, replaced with new low ESR caps, which have worse ESR. PSU still hisses.

You should rather check/replace small electrolytic capacitors. Especially on the primary side. Jamicon is Taiwanese.
 

replaced with new low ESR caps, which have worse ESR

Aren't those Chongex by chance?

[Armadillo]

I suggest if you can, you should take out the transformer and carry out a simplified Hi-Pot testing with a HV insulation tester.
And then with a ring tester.
if it fails, don't waste your time changing those capacitors.

PC power supplies [used], those < 500W are in the cheap, people just dump away literally speaking. That kind of power will be a big horse extra redundant power supply for your 486.
Of course I don't know whether this is the kind of practice in your country.
If it's 1000W and above, then is worth while to rescue it. That's my opinion of course.

If the transformer is bad, you should know what to do. Changing the capacitors won't resolve the problems.

[AndreyS]

You should rather check/replace small electrolytic capacitors. Especially on the primary side. Jamicon is Taiwanese.

I think, I'll give it a try and replace those caps with good quality ones next week

Aren't those Chongex by chance?

Almost guessed, two are Chong, one is Chengx

[ArthurDent]

The hiss you hear could be just the switching frequency used by the pulse width modulation and have nothing to do with the large caps or transformer leakage. The fact that the pitch changes with load indicates this could be the problem. If the frequency drops low enough it can be in the range of human hearing. If the output voltages are still correct it could also be a loose core on the converter transformer. If you can touch the core with a wooden ruler when it is making this sound and it changes pitch it probably is the two halves of the ferrite core material are loose.

[wraper]

Almost guessed, two are Chong, one is Chengx

Those are garbage. Dunno If those are even genuine branded to begin with. I've bought a few of those in local shop. Not only ESR was complete trash, worse than basically any general purpose cap, their voltage rating was fake too. They were too small for their ratings and testing them proved the suspicion. At 'rated' voltage they already had a few mA of leakage current. Add a few volts more and it was in hundred mA range.

[T3sl4co1l]

Hissing usually means the controller is fighting against an unstable and nonlinear loop.  The nonlinearity may or may not be a sign of faulty components, or an unintended part of the design; but it can definitely be exacerbated by bad caps.

Tim

[Armadillo]

For the purpose of discussion;

Let's recall the fluorescent lighting with the old ballast type starter. When the lighting begins to hum, we change the ballast and the hums disappeared. Yeah!, some will say change to electronic ballast. That will certainly work as well. [Edit: The exacerbation in this case is the Ballast itself].

Obviously, distinction must be made between humming and hissing noise. Hissing because the transformer is small, lighter in mass and the frequency is higher.

While other causes are likely, but when the transformer begin to hiss, it is very likely that it has been impaired, not forgetting that PC SMPS does come with some forms of Protection etc.

Not denying other possibilities, but the question then would be "Economics & Risks" of recapping without some forms of prior testing, in my opinion.

Above is only for discussion. Is OP Preference and Choice, obviously. 


 

[AndreyS]

The hiss you hear could be just the switching frequency used by the pulse width modulation

That was my initial thought too.
But now I am a bit confused considering all the responses all of you have written.

[Armadillo]

Ask yourself, so far of all the time and money you have spent changing the capacitors, have you found any objectionable ones and the problems resolved?

I read above it's counter productive to change worse ESR than original caps, yes? From big ones, you were led on to change smaller ones.

The question is BAD one. If you are following their episodes, than find the bad ones and make the day.

IMO, Changing caps must be supported by some technical reasons, like for example, start up problem, sluggish regulation, high ripples, output voltage out of specs. etc.... and yes, noise but as a secondary results of the high ripples, all pointing to measurable efforts other than blind changing. I know its a bit difficult to test the transformer but you can swop if there is equal spare.

Just my opinions.  Cheers.   
 

The hiss you hear could be just the switching frequency used by the pulse width modulation

That was my initial thought too.
But now I am a bit confused considering all the responses all of you have written.