Checking PWR_OK and PSU_On with a DMM

[quest for silence]

Ok, I need a quick'n'dirty recap how to properly do that.

With the PSU plugged into mains outlet, pins 15-16 not shorted, back PSU switch (if any, of course) set to "ON", with the DMM I should read on the 20 pin ATX connector from pin 9 (5VSB) 5V, from pin 14 (PSU On) something higher than 2.5V, and from pin 8 (PWR_OK) = ? (I don't recall)

Right?

Then, after shorting the pins 15-16 with the paperclip, the DMM should read from pin 9 still 5V, from pin 14 something less than 1V, and from pin 8 (PWR_OK) = ? (as above)

Right again? Or is this all wrong?

Well, thanks in advance for hints/help.

[quest for silence]

Jeez, in more than two hours no DMM man stumbled upon here... what a luck!

Well, rephrasing/amending/summarizing:

a) paperclip off: 5VSB =5V / PSU_ON > 2.0V / PWR_OK < 0.4V
b) paperclip on: 5VSB =5V / PSU_ON < 0.8V / PWR_OK > 2.4V

Can anyone confirm whether that's right or not? Thanks in advance!

[ridgid13579]

Jeez, in more than two hours no DMM man stumbled upon here... what a luck!

Well, rephrasing/amending/summarizing:

a) paperclip off: 5VSB =5V / PSU_ON > 2.0V / PWR_OK < 0.4V
b) paperclip on: 5VSB =5V / PSU_ON < 0.8V / PWR_OK > 2.4V

Can anyone confirm whether that's right or not? Thanks in advance!

Sorry, I was asleep still... Time zones...
Anyways, you could always read the atx spec :P

This is what I got (Tested with the CX-450M):

Paperclip off: PSU_ON is 2.5v, PWR_OK should be low (0v).
Paperclip on: PSU_ON is 0v, PWR_OK signal should be 5v, not 2.4v.

5vSB will always remain the same 5v, it's a separate rail (I confirmed it as well).
With the paperclip on, it should be zero. You're pulling it low by grounding it with a paperclip that is tied to ground, so I guess 0.8v is just an error margin?? (although close to 1v shouldn't be that high of an "error", what multimeter + probes are you using?).

[quest for silence]

Sorry, I was asleep still... Time zones...

Thank you ridgid, BTW I know north american guys were under the moon: I was thinking to my european comrades, first and foremost our The Mask (hey mate, how many PSUs did you checked before now? Countless I guess).

Anyways, you could always read the atx spec :P

If I were not illiterate, yes, I could.

BTW, that's where I took those values (just who knows whether I read the right pages or not!).

PWR_OK signal should be 5v, not 2.4v.

Yes, that's what I think I read in the book, >2.4V and <5V: I wrote >2.4V for the sake of brevity, sorry for misleading anyone.

I guess 0.8v is just an error margin?? (although close to 1v shouldn't be that high of an "error",

No, no, actually I didn't do anything atm: it's what's written on ATX specs (albeit maybe I took a wrong revision).

what multimeter + probes are you using?).

I'm going to use my cheap'n'trusty UNI-T UT139C with stock probes.

[ridgid13579]

Oh, I didn't even look at the spec :P. But I guess it makes sense..... Most things will consider 2.5v as high if they run on 5v or 3.3v...

I'm going to use my cheap'n'trusty UNI-T UT139C with stock probes.

Oh that doesn't look like a bad meter for it's price. I thought you were using some super sketchy ebay meter, but at this low voltage it doesn't really matter does it?

[quest for silence]

Oh that doesn't look like a bad meter for it's price.

Well, actually it's one of the more well regarded on EEVblog (among cheap ones, of course), and that's why I bought it (I thought that it were a safe buy).