PSU problems. Can someone interpret my voltage readings?

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View

I am having some problems with my PC. Here are the symptoms:

1) The PC comes on for a second, goes off, then comes on again. It
boots through POST and halts with a CMOS Checksum error where I have
to reconfigure the BIOS (I think this is a red herring). After BIOS
reconfig PC boots and works happily.

2) Once turned off after 1) the next time it is booted it will just go
off again after 1 second. This time the power switch does nothing. To
get round it I power off the PSU (or pull the lead) wait 20 seconds
for power to drain, reconnect and reboot. Sometimes it goes straight
into Windows, other times we go back to 1).

After reading some handy hints from w_tom I have started my
investigations. Here are my numbers:

Purple: 5.02v
Green: 5.00v**
Grey: 5.10v on power on
Red: 5.10v on power on
Orange: 3.38v on power on*
Yellow: 11.74v

*This is the anomaly I can see. It is too low. What would this low
voltage cause?
** Is Green too high?

Orange is 0.07v when no power is connected (?) and 0.19v when off but
power connected.

I also measured the V on the power switch jumper. This read 0.93v. Is
this too low?

I measuered the current by removing the AC cord, disconnect the power
connectors and attaching the meter leads to green and black. After
reconnecting the AC cord the reading was 0.14.

I tested the PSU with a PSU tester and all lights came back green for
all voltages.

Finally, a video speaks louder than words so here is one showing the
above problem:


Re: PSU problems. Can someone interpret my voltage readings?

Jimbob wrote:
Quoted text here. Click to load it

You can probably fix that by replacing the battery on the mainboard.

Quoted text here. Click to load it

That voltage is fine.  Here is a list of expected voltages for all 20
pins of the ATX power connector.  (24 pin power connectors still use the
same colors for the same voltages.)

Quoted text here. Click to load it


The value of taking voltage readings, on anything other than standby
power, when the PSU is off (or worse, unplugged) escapes me.  Someone
else will have to comment if those numbers are in any way meaningful.

Checking voltages can really only identify a problem with a PSU.  If the
voltages are good, the supply may still not be able to deliver the
current under load.  The simplest and most reliable way to check for a
bum supply is to substitute a known good supply in its place.

Re: PSU problems. Can someone interpret my voltage readings?

Quoted text here. Click to load it

  First, when red, oragne and yellow voltages are as posted, then
green wire voltage must not be as posted.  Two numbers should be
provided for each wire.  Voltage before power switch is pressed AND
numbers that describes voltages as power switch is pressed.

  For example, green wire voltage would be well above 2.0 volts when
power supply is off.  When power switch is pressed, voltage must drop
to well below 0.8 volts.  Green wire is the power supply controller
ordering power supply on.  It should not always be 5.1 volts.

   Second, voltages are in spec.  But curious is why 3.3 and 5 are
slightly high when 12 volts (yellow) is so low and marginal.  This
would imply power supply was not properly sized for the load - yellow
wire was not intended for so much 12 volt energy consumption.  Well
this is a marginal condition; probably would not explain your
shutdown.  Yellow wire is only 0.04 volts above what is absolute
minimum for the meter.  Keep this low number in mind while moving on
looking for something that would better explain your symptoms.

  Third, to better answer questions, required are numbers before power
switch is pressed AND what each wire does as power switch is pressed.
Based upon the description, I would expect gray wire to never rise
about 1 volt after power swtich is pressed.  Does it rise to 5.1 volts
after power switch is pressed?  Or is it always 5.1 volts?  Gray wire
is a power supply reporting status to the power supply controller.  It
should rise within two second after other voltages have stablized.

  If power supply controller does not see good on that "Power OK"
signal wire, then power supply controller powers off the supply -
typically within two seconds.  Sound familiar?   However, to know
more, required are voltage numbers for each wire both before and when
power switch is pressed.

Re: PSU problems. Can someone interpret my voltage readings?

On Fri, 25 Apr 2008 11:57:09 -0700 (PDT), Jimbob

Quoted text here. Click to load it

You left out very important bits of information - the
specifics about the hardware in this PC, including CPU,
motherboard make/model, PSU make/model/ratings.

Had the system previously ran properly and if so, what has
changed since then?

Quoted text here. Click to load it

We cannot conclude anything about this necessarily because
we dont' know what hardware you have.  Certain chipsets and
bios do this when there is a compatibility problem or
settings are changed.  If it is a bios issue, (and you are
confident it is stable once booting) you might seek a newer
bios.  CMOS checksum error can also be caused by a bad
battery, confirm the correct orientation of the clear CMOS
jumper and disconnect AC, pull the battery and check it's
voltage (Or put a new one in if all else fails), use the
clear CMOS jumper then put battery back and connect AC power
again.  It is expected you may have to enter bios and save
settings one more time, do not just enter bios and exit
without saving (even if you make no changes, "save" when

Quoted text here. Click to load it

This may still be a motherboard logical problem, though a
short or malfunctional PSU could be to blame.

Quoted text here. Click to load it

Orange should be 3.3V, it is not too low.  When are these
readings taken?  It would be important to check voltage
during the period of turn on initialization since that may
be when the PSU can't cope with the turn on surge.   Yellow
(12V) looks a little low at 11.75V but is still within spec,
if the PSU is otherwise working properly this voltage level
is not a problem... but it could also be a sign the PSU
doesn't have enough 12V current or is regulating from a 5V
feedback and won't respond properly to changes in 12V
current requirement.  Again having information about the
system specific hardware is a good place to start rather
than finish.

Quoted text here. Click to load it

I can't entirely rule out PSU from the symptoms but suspect
it is a bios bug.  What CPU are you using, what memory, and
what bios settings?  If it is a bios problem and there is no
newer bios, it might be possible to try different manual
bios settings as a temporary solution - though depending on
what settings are needed, the system might not run at the
proper default speed(s) in the interim.

In some instances there are larger groups of fellow owners
of particular motherboards at web forums focused on such
things.  From a brief look at the youtube video it appears
to be a Foxconn motherboard.  Unfortunately I don't know
whether there are any Foxconn web forums but Google might
find one.

If this is a newly purchased motherboard and you have an
opportunity to return it to the seller for a refund, that
might be the best strategy - before time runs out on your
chance to do so (assuming nothing else we cover here helps).

Site Timeline