Some PSU Rails too low, others are fine

Thread Starter

Daan Lageschaar

Joined Jul 8, 2019
45
Hi there!

EDIT: Forgot to mention that the +5V rail is fine as well at around 5.2V

I have a BSS FDS-336 audio DSP that was hooked up to a power generator with unclean power. The generators voltage would fluctuate down between 206-260VAC all the time. This must've damaged something in the power supply section because now the DSP occasionally displays "power failure" on screen, which in turn means the DSP doesn't turn on the output relays.

The DSP has an Switch mode psu with multiple voltage rails on the secondary side. -18V, +18V, +8V and +5V

The 2 rails that seem to be affected/damaged are the +18V and -18V rails, from which +15V and - 15V are made using a TIP32A and a TIP31A.
The -18V rail is around -14.2V
The +18V rail is around 13.5V

now the weird thing is that the +8V rail is just fine, while it's on the same transformer secondary, so it can't be a issue with the PWM controller chip or something similar.

I've tried multiple things to fix this issue, a new PWM chip (just in case), A new switching MosFET, I tried replacing the rectification diodes on the secondary side with some 1n5408's I had laying around, but those got incredably hot and started smoking.

I've also attached the schematic for the PSU to clear things up a little bit more
The voltage rails on the secondary side are, from top to bottom:
+15V
+8V
-15V
+5V
 

Attachments

Last edited:

AlbertHall

Joined Jun 4, 2014
12,188
Are the loads on the outputs what the supply is expecting?
For instance, The supply regulates only the +5V supply so if this has a very light load or no load the other outputs will be lower than expected, or if the loads on the ±18V are higher than usual then their voltages will be lower.
 

Thread Starter

Daan Lageschaar

Joined Jul 8, 2019
45
Are the loads on the outputs what the supply is expecting?
For instance, The supply regulates only the +5V supply so if this has a very light load or no load the other outputs will be lower than expected, or if the loads on the ±18V are higher than usual then their voltages will be lower.
The loads on the +/- 18V rails seem to be normal to me, around 160mA for both of them
I forgot to mention in the original post BTW that the +5V rail is fine as well at around 5.2V
 
Last edited:

Thread Starter

Daan Lageschaar

Joined Jul 8, 2019
45
I wonder if you have tested (for ESR) or replaced all of the caps on the secondary side? Not just on the +/- 18V rails?
I'm not really sure how to measure the ESR, since I don't have any equipment for that
And I did indeed only replace the caps on the +/-18V rails since those 2 were too low. If another rails caps are bad then that rail would be to low, right?
 

Denis6482

Joined Sep 20, 2022
7
If I've understood correctly the 5V rail is used for regulation in this case and although the 5V rail Voltage appears correct there might be ripple which can affect the non-regulated outputs. If you don't have an ESR meter a 'scope might help here.
 

Thread Starter

Daan Lageschaar

Joined Jul 8, 2019
45
If I've understood correctly the 5V rail is used for regulation in this case and although the 5V rail Voltage appears correct there might be ripple which can affect the non-regulated outputs. If you don't have an ESR meter a 'scope might help here.
Ah good suggestion, didn't think of that yet. I'll try replacing the 5V rail caps when I get home. I only have some 820uF 25V caps left though, that won't be too big of a problem I hope?
 

Denis6482

Joined Sep 20, 2022
7
I don't think 25V replecements would be suitable, the ESR of an electrolytic increases with it's Voltage rating. That's why the 5V rail of PSU's is often fitted with 6.3V caps. They're smaller, cheaper and perform better than higher rated caps.
 

Thread Starter

Daan Lageschaar

Joined Jul 8, 2019
45
I don't think 25V replecements would be suitable, the ESR of an electrolytic increases with it's Voltage rating. That's why the 5V rail of PSU's is often fitted with 6.3V caps. They're smaller, cheaper and perform better than higher rated caps.
The first filter cap on the 5V rail is allready a 25V one, the second one is 16V, so I'll try to get a suitable 16V replacement. Maybe I'll replace the first filter cap with 16V as well if I can find a suitable one as well
 

Denis6482

Joined Sep 20, 2022
7
Sounds like a plan. Perhaps use 10V if you are buying. And sorry to be repetitive you need low ESR, I use Rubycon Nichicon and other good brands.
 

Thread Starter

Daan Lageschaar

Joined Jul 8, 2019
45
Sounds like a plan. Perhaps use 10V if you are buying. And sorry to be repetitive you need low ESR, I use Rubycon Nichicon and other good brands.
Replaced ALL caps on the secondary side with 25V 820uF low ESR Rubycons I still had laying around at home, but that didn't make any difference unfortunately
 

Thread Starter

Daan Lageschaar

Joined Jul 8, 2019
45
I also replaced the switching diodes, and that also didn't help. So I'm gonna go ahead and guess it's just some broken transformer windings or something like that
I also measured all voltages using a scope, and they seem perfectly stable, albeit a bit too low on the +/-18V rails
 
Last edited:

Denis6482

Joined Sep 20, 2022
7
Daan, the rectifiers on the secondary side, you tried 1Nxxxx replacements which are of course too slow. What have you got fitted there now and are their temperatures normal?

If the PSU has been stressed, is the pair of 1R resistors on the primary, switching FET, Source to Ground within tolerance?

I'm not suspecting the transformer here. I assume you've checked for dry joints etc.

Also, bottom left of the schematic, PSU sync, does your 'scope show activity there?
 
Last edited:

Thread Starter

Daan Lageschaar

Joined Jul 8, 2019
45
Daan, the rectifiers on the secondary side, you tried 1Nxxxx replacements which are of course too slow. What have you got fitted there now and are their temperatures normal?

If the PSU has been stressed, is the pair of 1R resistors on the primary, switching FET, Source to Ground within tolerance?

I'm not suspecting the transformer here. I assume you've checked for dry joints etc.

Also, bottom left of the schematic, PSU sync, does your 'scope show activity there?
Right now the original diodes are back in, but other ones also didn't help. I'll check the resistors and PSUSYNC once I get home
 
Top