Re: Really stumped: PS inoperative

I've benefited from your and W_tom's advice and have made some
progress. The bottom line is that the 12V bus reads correctly if the MB
is removed from the case; not if it is inserted. I probably have a short
somewhere, perhaps something between the MB and its back plate. The next
step will be to remove it from that back place, but that brings up a
question: Is the ground provided by the back main power connector
sufficient to ensure system ground for testing purposes?

In going over the steps to isolate the cause of an undervoltate in one
of the three power busses. I've had to expand or edit your helpful
advice a bit, and perhaps easiest would be for me to provide the result
and have you spot where I go wrong:

1. To isolate the problem to the MB+PSU, remove all cards except the
video card from the MB [why not the video card?], CPU power connector,
all front panel connectors and optionally the power switch, and all PSU
power going to any peripheral drives. Leave connected only the MB main
power connector. Briefly short the PWR_On (green, pin 14) to ground, or
the pins on the front panel header that would connect to the power
switch, or push the power switch if it is connected [strange, this test
condition seems to have made pressing the power switch unnecessary, for
the 12V bus went right up an acceptable value as soon as I turn the PSU
on. Given that I have a jury-rigged power switch, I'll have to make sure
it is not accidentally grounded]. Measure the voltage on the three rail
wires: orange, red and yellow, which should rise to at least +3.23V,
+4.87V, and +11.7V respecively within a second.

2. If these voltages fail to rise properly, remove the MB and PSU to see
if powering up the MB external to the case corrects the problem. [I find
here that since this removal did correct the problem, to determine that
the MB was the culprit, I had to resort to a PSU swap. Is there a better
way?]

3. Or if these voltages do rise properly, then we are in a position to
isolate the problem. Restore the CPU power connection to see if it rises
to about 12V after the power switch is pressed. If there's a CPU short,
there should be only a very slight rise in voltage at the CPU power
connector when the power switch is pushed, but no rise close to the
short itself [I'm not sure quite how to implement this as far as
identifying and accessing test points. My MB, although new, was built
when the electrolytic problem first came came up, and so of course,
inspecting them was an obvious thing to do. I'm not confident that part
replacement is all so easy on a multi-level printed board.]

4) Restore the cards back to the MB and repeat the voltage checks
above. If the voltages are OK, presumably the problem is not that of any
particular card.

5) Reconnect the peripherals one by one, checking voltages each time.

--

Haines Brown, KB1GRM



.

Relevant Pages

  • Re: Power up problem
    ... black/green pins on the 20 pin mb connector and them press my start button.. ... Under mormal shutdown conditions in which I power off it restarts fine.. ... Also useful is to measure voltages on green and gray ... numbers and how those voltages responded as power switch is pressed. ...
    (alt.comp.hardware.pc-homebuilt)
  • Attn: Paul/WTomRe: Power up problem
    ... black/green pins on the 20 pin mb connector and them press my start button. ... Under mormal shutdown conditions in which I power off it restarts fine.. ... Also useful is to measure voltages on green and gray ... numbers and how those voltages responded as power switch is pressed. ...
    (alt.comp.hardware.pc-homebuilt)
  • Re: what to do next?
    ... you still don't know even if the power supply ... Procedure is in "When your computer dies without warning....." ... Important are those numbers both before and when power switch is ... plugged in), from backprobing the MB power connector, the voltages ...
    (sci.electronics.repair)
  • Re: Power issue follow-up question: UPS?
    ... But if we look at what happens when there is a power outage / power flicker ... charts for voltages that are ... Same charts also define high voltages that 120 volt electronics must ... How do we limit those spikes so that internal appliance protection is ...
    (alt.guitar.bass)
  • Re: Unstable A7N8X-DLX 2.0 system, Power Supply was culprit
    ... I began to suspect my vid card, vid card drivers or some other ... quite surprised to see Speedfan's graph of the output voltages. ... 5v rail was down to about 4.65-4.70v at startup, ... I've measured the power on my A7N8X-E Deluxe, with a clamp-on DC ammeter, ...
    (alt.comp.periphs.mainboard.asus)
Loading