PSU Failing or what?

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

Here are some readings...

VCOREA=1.70V
VINRO=1.70V
+3.3v = 3.44v
+5v=5.08V
+12V=11.67V
-12V=-11.88V
-5V=-4.62V

What does this mean? a failing psu?
also these values fluctuate wildly especially the +/-5 and the neg12V
one does.

My comp also does a reboot for unknown reasons on occasions.
thanks,

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

dnc_tn@yahoo.com wrote:

> Here are some readings...
>
> VCOREA=1.70V
> VINRO=1.70V
> +3.3v = 3.44v
> +5v=5.08V
> +12V=11.67V
> -12V=-11.88V
> -5V=-4.62V
>
> What does this mean? a failing psu?

What does what mean? The numbers?

What did you expect them to be, because they look ok.

> also these values fluctuate wildly especially the +/-5 and the neg12V
> one does.

Fluctuating numbers is not ok. What are you reading the numbers with?

>
> My comp also does a reboot for unknown reasons on occasions.
> thanks,

If the PSU really is "fluctuating wildly" then that could cause it, but
then I don't know what you consider 'wildly' to be or whether the monitor
program is doing the fluctuation (misreading).

And, of course, other things can cause spurious reboots too.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

On Tue, 11 Jan 2005 19:48:35 -0500, "dnc_tn@yahoo.com"
<DNC_TN@> wrote:

>Here are some readings...
>
>VCOREA=1.70V
>VINRO=1.70V
>+3.3v = 3.44v
>+5v=5.08V
>+12V=11.67V
>-12V=-11.88V
>-5V=-4.62V
>
>What does this mean? a failing psu?
>also these values fluctuate wildly especially the +/-5 and the neg12V
>one does.
>
>My comp also does a reboot for unknown reasons on occasions.
>thanks,


+12V reading looks right for a motherboard sensor reading,
but if it were from a multimeter reading at power connector,
more likely then a sign of insufficient 12V rail capacity.

It's not surprising that negative values are fluctuating if
unloaded, which is probably the case.

What CPU do you have that has 1.7V Vcore? Perhaps it's
overclocked (therefore overvolted intentionally)? Sometimes
a mobo's vcore will be a little high on a CPU with 1.65V
default vcore, but generally not a whole .05V higher. Even
so, too high a vcore shouldn't be rebooting system unless
CPU is overheating.

Rebooting could simply be windows errors, disable the
windows reboot on error setting (Google searching will find
details on doing that). If it seems more related to power
or temp, more closely monitor status of those readings while
reproducing events causing reboot.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

The 3.3 volts is marginally high. Just under the limit of
being too high and therefore suspect. 12 volts is just above
the limit. Again, you did not say what measured voltage -
what hardware. Therefore these conclusions are still
invalid. But from that 12 volt number and if using a 3.5
digit multimeter, then I would investigate further for too
much ripple voltage on that supply - a slowly failing internal
part.

Again, it sounds like you thought motherboard readings were
reliable. Wrong. Requirement of a 3.5 digit multimeter is
the cheapest method possible to measure voltage with any
reliability. The 3.5 digit meter then calibrates motherboard
voltage monitor.

-5 also does not look good. But -5 typically is not used
for anything important. Where is +5 VSB? Numbers for that
important voltage were not provided. Get the meter. Some
examples of how to measure important voltages: "Computer
doesnt start at all" in alt.comp.hardware on 10 Jan 2004
at
http://tinyurl.com/2t69q and
"I think my power supply is dead" in alt.comp.hardware on 5
Feb 2004 at http://tinyurl.com/yvbw9

"dnc_tn@yahoo.com" wrote:
> Here are some readings...
>
> VCOREA=1.70V
> VINRO=1.70V
> +3.3v = 3.44v
> +5v=5.08V
> +12V=11.67V
> -12V=-11.88V
> -5V=-4.62V
>
> What does this mean? a failing psu?
> also these values fluctuate wildly especially the +/-5 and the
> neg12V one does.
>
> My comp also does a reboot for unknown reasons on occasions.
> thanks,

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

w_tom wrote:

> The 3.3 volts is marginally high. Just under the limit of
> being too high and therefore suspect. 12 volts is just above
> the limit.

His 12 volt rail, at only 2.75% down, is well within limits, and not
'above' them.

> Again, you did not say what measured voltage -
> what hardware. Therefore these conclusions are still
> invalid. But from that 12 volt number and if using a 3.5
> digit multimeter, then I would investigate further for too
> much ripple voltage on that supply - a slowly failing internal
> part.
>
> Again, it sounds like you thought motherboard readings were
> reliable. Wrong. Requirement of a 3.5 digit multimeter is
> the cheapest method possible to measure voltage with any
> reliability. The 3.5 digit meter then calibrates motherboard
> voltage monitor.
>
> -5 also does not look good. But -5 typically is not used
> for anything important. Where is +5 VSB? Numbers for that
> important voltage were not provided. Get the meter. Some
> examples of how to measure important voltages: "Computer
> doesnt start at all" in alt.comp.hardware on 10 Jan 2004
> at
> http://tinyurl.com/2t69q and
> "I think my power supply is dead" in alt.comp.hardware on 5
> Feb 2004 at http://tinyurl.com/yvbw9
>
> "dnc_tn@yahoo.com" wrote:
>
>>Here are some readings...
>>
>>VCOREA=1.70V
>>VINRO=1.70V
>>+3.3v = 3.44v
>>+5v=5.08V
>>+12V=11.67V
>>-12V=-11.88V
>>-5V=-4.62V
>>
>>What does this mean? a failing psu?
>>also these values fluctuate wildly especially the +/-5 and the
>>neg12V one does.
>>
>>My comp also does a reboot for unknown reasons on occasions.
>>thanks,

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

Limits for 12 volt supply: 11.4 V to 12.6 V
His voltage: 11.67V. Therefore as previously posted:
> 12 volts is just above the limit.
When taking voltage with a 3.5 digit multimeter, the readings
must be in the upper three quarters of those limits. Voltages
in the lowest quarter may be due to excessive ripple voltage.
This is a failure that has not yet happened. Multimeters
cannot measure ripple voltages. Therefore a voltage in the
lower quarter of those limits suggests further study may be
required using other equipment.

His 12 volt readings may be an indication of excessive
ripple voltage that will only get worse in the future.

But more important, we don't know what was used to measure
these voltages. If measured using the motherboard monitor,
then the readings are not reliable.

He has three problems. 3.3 is suspiciously high. 12 volts
might be questionable. Source of these numbers is highly
suspect. It may be a good supply. But we don't know yet from
numbers as provided.

David Maynard wrote:
> His 12 volt rail, at only 2.75% down, is well within limits, and
> not 'above' them.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

On Thu, 13 Jan 2005 14:36:18 -0500, w_tom
<w_tom1@hotmail.com> wrote:

> Limits for 12 volt supply: 11.4 V to 12.6 V
> His voltage: 11.67V. Therefore as previously posted:
>> 12 volts is just above the limit.
>When taking voltage with a 3.5 digit multimeter, the readings
>must be in the upper three quarters of those limits. Voltages
>in the lowest quarter may be due to excessive ripple voltage.
>This is a failure that has not yet happened. Multimeters
>cannot measure ripple voltages. Therefore a voltage in the
>lower quarter of those limits suggests further study may be
>required using other equipment.
>
> His 12 volt readings may be an indication of excessive
>ripple voltage that will only get worse in the future.
>
> But more important, we don't know what was used to measure
>these voltages. If measured using the motherboard monitor,
>then the readings are not reliable.
>
> He has three problems. 3.3 is suspiciously high. 12 volts
>might be questionable. Source of these numbers is highly
>suspect. It may be a good supply. But we don't know yet from
>numbers as provided.
>
>David Maynard wrote:
>> His 12 volt rail, at only 2.75% down, is well within limits, and
>> not 'above' them.


Primarily, OP omitted basic critical details of the
motherboard and power supply. We have no reference of power
distribution for the system.

I agree that 11.67V is in-spec BUT that a motherboard
reading cannot be relied upon, in general most motherboards
do read a lower than actual 12V level when PSU output is
spot-on 12.0V. Multimeter readings at power supply
connector to motherboard are crucial for 12V reading.

Having written that, if the multimeter reading is at 11.67,
it is a sign that the power supply is of insufficient
capacity, else not appropriate for the system (typically
using a system with 12V power for CPU but not an ATX v2.03
power supply, so the PSU has little to no 12V feedback
weighting for regulation, rather it's regulating based
mostly or entirely upon 5V reading, which would coincide
with voltage readings reported. Even so, CPU VRM circuit
will remain stable far below 11.67V, most would be OK below
10V, but this tells nothing of the power supply's other
rails.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

When using a digital voltmeter, voltages in the bottom
quarter of voltage limits are undefined. A ternary world
where we have good, bad, and undefined conditions. It may be
good or it could also be bad - therefore undefined.

That voltage reading is 11.67. Number is within voltage
limits but is in bottom quarter. Now that could be 11.67
volts with a ripple voltage of 50 mv. Or that could be a
voltage that varies between 11.3 and 11.7 because the 12 volt
output has too much ripple voltage - 400 millivolts. Voltage
is still undefined. A meter often tends to read the upper
voltage of that ripple wave. Therefore that 11.67 volts
really might be 11.5 with excessive (out of spec) ripple
voltage. IOW voltage dropping repeatedly to 11.3 volts.

IOW we don't know if that 12 volts is good because at 11.67
volts, it is in the bottom 1/4 of those limits AND because
multimeters don't always read RMS; cannot read higher
frequency ripple voltages.

Again, that 3.3 volts is in limits but suspiciously high.
And that 12 volts is also in a region of suspicion. Yes the
system is often very tolerant of bad 12 V DC. But a system is
often less tolerant of high ripple voltages. 11.67 may be a
symptom of excessive ripple voltage.

kony wrote:
> On Thu, 13 Jan 2005 14:36:18 -0500, w_tom
> <w_tom1@hotmail.com> wrote:
>> Limits for 12 volt supply: 11.4 V to 12.6 V
>> His voltage: 11.67V. Therefore as previously posted:
>>> 12 volts is just above the limit.
>>When taking voltage with a 3.5 digit multimeter, the readings
>>must be in the upper three quarters of those limits. Voltages
>>in the lowest quarter may be due to excessive ripple voltage.
>>This is a failure that has not yet happened. Multimeters
>>cannot measure ripple voltages. Therefore a voltage in the
>>lower quarter of those limits suggests further study may be
>>required using other equipment.
>> ...
>
> Primarily, OP omitted basic critical details of the
> motherboard and power supply. We have no reference of power
> distribution for the system.
>
> I agree that 11.67V is in-spec BUT that a motherboard
> reading cannot be relied upon, in general most motherboards
> do read a lower than actual 12V level when PSU output is
> spot-on 12.0V. Multimeter readings at power supply
> connector to motherboard are crucial for 12V reading.
>
> Having written that, if the multimeter reading is at 11.67,
> it is a sign that the power supply is of insufficient
> capacity, else not appropriate for the system (typically
> using a system with 12V power for CPU but not an ATX v2.03
> power supply, so the PSU has little to no 12V feedback
> weighting for regulation, rather it's regulating based
> mostly or entirely upon 5V reading, which would coincide
> with voltage readings reported. Even so, CPU VRM circuit
> will remain stable far below 11.67V, most would be OK below
> 10V, but this tells nothing of the power supply's other
> rails.