Delta DPS 320 EB missing power good signal

[ultrasmurf]

Hi all,

I'm troubleshooting this delta DPS 320 EB. Problem : PG Signal is at 0V. Leaky capacitor has all been replaced and could not help to think that the problem lies in the DNA1002 chip which from what I read is a custom IC by delta.

I'm using the following schematic as it looks quite similar (not the same at all in terms of component value but more or less similar) : https://cdn.badcaps-static.com/pdfs/...2cdd095645.pdf

Measurement at the output :
5V : 5.15
12V : 11.7
-12V : -12.08
3.3V : 3.37

under load 5v drop to 5v and 12v line drop to about 11.45v. Values doesn't look too good but should still be within ATX spec (not sure about DNA1002 expected value).

Measurement on DNA1002 pin shows similar number like the one on the simonwyss website, except pin 3 (PG : 0v expected 5v), pin 8 (Unknown : 2.6-2.7v expected 4.25v), pin 9 (5Vin : 5.15) pin 10 (12Vin : 11.7V), pin 14 (unknown : 2.6v expected 2.7v). Expected value is based on symonwyss number. All other pin are pretty much spot on with his measurement.

Adding filtering capacitor on 5V and 12V lines does not help.

Any ideas where could the problem lies ?

Thanks for the help...

[momaka]

Re: Delta DPS 320 EB missing power good signal

Hi all,

I'm troubleshooting this delta DPS 320 EB. Problem : PG Signal is at 0V. Leaky capacitor has all been replaced and could not help to think that the problem lies in the DNA1002 chip which from what I read is a custom IC by delta.

Can you elaborate on that? Which capacitors were replaced and which weren't? The output capacitors (3.3V/5V/12V/-12V rails)? - All of them? What about the small capacitors near the control board(s)? Replaced those too?

If there are any bad cap brands still left in there, there is no point to try and troubleshoot with measuring voltages and whatnot - you need to replace all capacitors (except the big primary caps). Otherwise, there is not telling if one of those small buggers may be messing up the whole circuit or not.

Also, pictures of your power supply would be helpful.

[ultrasmurf]

Re: Delta DPS 320 EB missing power good signal

Output capacitor are all tested good (surprisingly...). Primary cap is also tested good. Bad capacitor are mostly on the control path and bypass cap for the IC. 47uf and 4.7uf type.
Capacitor testing is using oscilloscope and function gen set at 125khz square wave 250mVp for onboard testing, if it shows some high ESR then taken out and replaced ( for a 250mV signal i define it as questionable if it still shows a 5mV signal after connected to the cap ). Did a test adding 2200uF parallel cap on all the output cap and it does not fix the issue.

Output voltage was observed with oscilloscope and ripple is well within ATX spec even without adding parallel cap to the output capacitor.

Will try to add picture soon...

[ultrasmurf]

Re: Delta DPS 320 EB missing power good signal

Picture of the PSU. Replaced capacitor is highlighted with red square. Additional note : This is PSU for B2600 HP Workstation, after replacing these bad capacitor workstation boots up and running for 3 days straight without issue. I also have a second workstation that I have replaced it with normal ATX power supply (swapping the pin around to match B2600 pin) and that is also running fine.

Next plan is to blindly replace every single electrolytic cap for both secondary and primary side to see if that brings the PG signal back on ( saw other thread reply encountering same issue on different psu ). Just that it doesn't make sense to me as output signal shows good ripple in the scope...

Few other additional note :

I can hear a faint tweet tweet tweet sounds roughly about 1 or 2 in a second that changes slightly depending on the load I introduced in the 12v or 5v output. Couldn't really pin point where it is coming from... I don't really have the guts to put my ear too close to a live PSU... who knows when those cap or transistor decide to blow themself up.

Adding a 12v100W halogen lamp load on the 3.3V line will make the PSU refuse to turn on ( a mere 7.5W load at 3.3V?), adding these load when PSU is running doesn't cause any issue.

[momaka]

Re: Delta DPS 320 EB missing power good signal

Bad capacitor are mostly on the control path and bypass cap for the IC. 47uf and 4.7uf type.

As I thought.
These small caps can cause lots of issues like this.

after replacing these bad capacitor workstation boots up and running for 3 days straight without issue.
...
Next plan is to blindly replace every single electrolytic cap for both secondary and primary side to see if that brings the PG signal back on ( saw other thread reply encountering same issue on different psu ). Just that it doesn't make sense to me as output signal shows good ripple in the scope...

It's probably the small caps if loss of PG signal is all there is wrong with the PSU.

Don't just replace blindly - replace blindly every little capacitor. The big ones, not that important for now (if ripple is within spec).

That said, try to use good quality low ESR capacitors rated for PSU use (and 105C). Don't use 85C general purpose stuff - that's no good, and you will be doing this recap again in the near future, because they won't last long in there.

Adding a 12v100W halogen lamp load on the 3.3V line will make the PSU refuse to turn on ( a mere 7.5W load at 3.3V?), adding these load when PSU is running doesn't cause any issue.

Halogen lamps have a very low resistance when cold - almost like a short-circuit. With the lamp connected before the PSU is turned ON, the PSU has to supply a large surge current to the halogen lamp and also its own output capacitors. This alone can be enough to trip the short-circuit, over-current, or under-voltage protections. Once the PSU is running and the output caps are charged up, they can provide that large surge current when the halogen lamp is cold, so that's why the PSU doesn't shut down once it's running.