1
\$\begingroup\$

I want to measure the power good time of ATX PSU using an oscilloscope.

As far as I know, normal time for power good is between 100ms to 500ms. and it is measured according to the picture below, starting from PS_ON command.

atx power supply timing diagram

So I measured PS_ON and power good signals using oscilloscope and got the following image [I forgot to get the original piscture, so I drew it]

measured signals on oscilloscope

I think the power good should the time between PS_ON falling edge and PG rising edge. which is 180ms. But when i connected power supply to an ATX power supply tester (i borrowed cheap one with no brand name), it displays 280-320ms as PG.

Which measurement is correct. mine, or power supply tester?

Any advice to correct measurement is appreciated.

Thanks.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ They both might be correct. You don't say how the tester loads the power supply, and you don't say how you load the power supply. Different loads get a different result. \$\endgroup\$
    – Justme
    Commented Dec 5, 2023 at 20:36

2 Answers 2

1
\$\begingroup\$

Both your measurement and the tester may be correct.

And actually your numbers for pass criteria are a bit wrong.

T1 is specified less than 500ms.

T2 is specified more than 100ms and less than 500ms.

T4 is specified less than 10ms.

So reading the diagram you posted, the absolute minimum allowed time from driving PS_ON# low to PG high is 100ms like you say, but even T2 is defined 0.1ms to 20ms.

Then for slowest possible supply, the PG is allowed to stay low 1000ms before it starts rising, and then must have risen within 1010ms.

The rate how fast all the outputs are stable depends on if you start the PSU with no or full load resistively, or with no or full load capacitively.

The power supply tester may have resistors and capacitors as load. How you did your test is unknown, but I bet you did not apply the exact same load as the tester. Different loads produce different time between PS_ON# and PG.

Since you have the tester, you can use the oscilloscope to verify if the tester measures the time properly, by measurint PS_ON# and PG with scope and tester.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks for your time to answer. Yes. you are right. I tested PSU at NO load state. I found and read the entire "ATX12V Power Supply Design Guide". So I need to verify the tester with simultaneous measurement. and test which timing parts are taken into account for displayed PG. So, apart from that, I can verify the state of PSU using oscilloscope considering T1, T3 and T4 (though the later may be negligible). am I correct? Thanks. \$\endgroup\$
    – Ali
    Commented Dec 6, 2023 at 5:01
1
\$\begingroup\$

If you trust the scope, I would use that then a power supply tester which may not have good software especially if it is cheap. If you put in a known source into the scope (like 60 or 50Hz) you could verify your scope, but generally scopes keep time better than milliseconds.

\$\endgroup\$
2
  • \$\begingroup\$ thanks ao much for your answer. So is my procedure to mesure, correct? and, Yes, I put in 50Hz and i trust my scope. \$\endgroup\$
    – Ali
    Commented Dec 5, 2023 at 19:47
  • 1
    \$\begingroup\$ Thanks, please remember to use the upvoting/answer system. Yes the procedure is correct \$\endgroup\$
    – Voltage Spike
    Commented Dec 5, 2023 at 19:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.