PSU voltages?

  • Thread starter Thread starter Hank
  • Start date Start date
H

Hank

Hey, I was looking at my PSU voltages in more than one Windows utility and
they all read about the same. The -12 reading seems to be a bit out of whack
and thought I would ask someone about it that knows more about PSUs than
me - That list would include just about everybody.

My voltages are:
+3.3 = +3.33
+5 = +5.08
+12 = +11.86
-12 = -13.43 and varies to -13.52
-5 = -5.30

System runs solid but is there anything to be concerned about???
Thanks a bunch!

Hank
 
Hey, I was looking at my PSU voltages in more than one Windows utility and
they all read about the same. The -12 reading seems to be a bit out of whack
and thought I would ask someone about it that knows more about PSUs than
me - That list would include just about everybody.

My voltages are:
+3.3 = +3.33
+5 = +5.08
+12 = +11.86
-12 = -13.43 and varies to -13.52
-5 = -5.30

System runs solid but is there anything to be concerned about???
Thanks a bunch!
Click to expand...

I would not worry about the voltages. The -12 is not used for much any
more. The voltages may not be that accurate anyway. You need to verify
them with a good voltmeter first.
 
Hank said:
Hey, I was looking at my PSU voltages in more than one Windows utility and
they all read about the same. The -12 reading seems to be a bit out of
whack and thought I would ask someone about it that knows more about PSUs
than me - That list would include just about everybody.

My voltages are:
+3.3 = +3.33
+5 = +5.08
+12 = +11.86
-12 = -13.43 and varies to -13.52
-5 = -5.30

System runs solid but is there anything to be concerned about???
Thanks a bunch!

Hank
Click to expand...

Like Ralph said, don't trust these values. If you only knew how these
numbers are created, you'd agree. Computerchips cannot measure voltages
higher then 4 or 5 volts or so because those measuringchips only run on
aprox 5 volts. The correct value is calculated by a formula. The same is
true for multimeters, but I think multimeters are a bit more precisly
calibrated then the software mainboard manufacturers use.

There is one problem. The -12V pin is only found in the ATX powerplug, so
you'll have to unplug the powersupply, and preferably have a special device
to properly load every channel and then measure the voltage. This can be
somewhat difficult to accomplish...

Halfgaar
 
Like Ralph said, don't trust these values. If you only knew how these
numbers are created, you'd agree. Computerchips cannot measure voltages
higher then 4 or 5 volts or so because those measuringchips only run on
aprox 5 volts. The correct value is calculated by a formula. The same is
true for multimeters, but I think multimeters are a bit more precisly
calibrated then the software mainboard manufacturers use.

There is one problem. The -12V pin is only found in the ATX powerplug, so
you'll have to unplug the powersupply, and preferably have a special device
to properly load every channel and then measure the voltage. This can be
somewhat difficult to accomplish...

Halfgaar
Click to expand...

-12V may still be off if there's no load on it, which isn't a problem.
Testing can be done by (carefully) inserting the probe into the back
of the ATX connector while the system is still powered, though again
if the -12V isn't loaded the voltage will still be off with no
resulting problems.


Dave
 
kony said:
-12V may still be off if there's no load on it, which isn't a problem.
Testing can be done by (carefully) inserting the probe into the back
of the ATX connector while the system is still powered, though again
if the -12V isn't loaded the voltage will still be off with no
resulting problems.
Click to expand...

I use a pair of sharp DVM probes, and just pierce through the wire
insulation just above the MB connector. Upon removing the probe, the
insulation almost closes over the small hole.
 
ric said:
I use a pair of sharp DVM probes, and just pierce through the wire
insulation just above the MB connector. Upon removing the probe, the
insulation almost closes over the small hole.
Click to expand...

Thanks guys, please excuse my ignorance here. Its been a long time since I
took that one semester of electronics way back in high school. I'm following
you about how to do the physical part of testing with the DVM, and I'm sure
I can figure out which settings to use and all, but you guys mention it may
or may not be under load. Now hold on to your hats here's the really dumb
question - How will I know if it is in fact under load or not?

Thanks a bunch guys!

Hank
 
Don't worry about the -12 volts. Even if under load, the
load would not care much about a voltage variation. The
critical voltages are 3.3, 5, and +12 volts. Those are the
voltages that must be in spec and that can create intermittent
failures if not.

-12 volts would really only be used by something that
required such voltage extremes such as the RS-232 serial
port. If any pins on that serial port are below -5 volts,
then you have identified something that takes from the -12
volts. Suppose the -12 becomes -11. RS-232 does not care
since even -5 volts would still be sufficient to drive that
RS-232 output.

-12 v might also be used by some device that uses
operational amplifiers. IOW the peripheral would be an analog
device. Those would be peripherals that might load -12
volts. Few peripherals take from the -12 volts which is why
many power supplies use a single chip, 1 amp voltage regulator
to create this voltage as if it was a separate power supply.
 
Thanks guys, please excuse my ignorance here. Its been a long time since I
took that one semester of electronics way back in high school. I'm following
you about how to do the physical part of testing with the DVM, and I'm sure
I can figure out which settings to use and all, but you guys mention it may
or may not be under load. Now hold on to your hats here's the really dumb
question - How will I know if it is in fact under load or not?
Click to expand...

There used be be cards like audio cards with onboard amps, which could
make use of -12V, more commonly it was used by serial ports, but IIRC
those now convert from 5V, so -12V isn't needed in today's PCs.

The surest way to check would be pulling the lead from the power
supply plug and checking current flow by connecting a meter in series
to the motherboard plug, but it's really not necessary, there is no
problem resulting from your voltage reading. I've used power supplies
that don't even have -12V or -5V rails without issue, though some
motherboard's BIOS will show an alert at POST due to it.


Dave
 
w_tom said:
-12 volts would really only be used by something that
required such voltage extremes such as the RS-232 serial port.
Click to expand...

I have some really old Socket 7 mobos that don't use the -12V rail for
RS-232 but instead have a chip with a couple of surface mount
resistors and capacitors to convert +5V to +12V and -12V. This seems
silly to me, especially since the capacitors on these particular mobos
have reliably exploded when somebody disconnected or plugged in an
RS-232 peripheral or even a parallel printer with the computer is
running. The capacitors explode so violently that there aren't any
bits rattling around, and the solder pads become nice and clean. At
first I didn't understand what failed because the RS-232 ports would
always measure OK with a voltage meter and test fine with a loopback
plug, but it turned out that the voltages sagged only at higher baud
rates.
 
kony said:
There used be be cards like audio cards with onboard amps, which could
make use of -12V, more commonly it was used by serial ports, but IIRC
those now convert from 5V, so -12V isn't needed in today's PCs.

The surest way to check would be pulling the lead from the power
supply plug and checking current flow by connecting a meter in series
to the motherboard plug, but it's really not necessary, there is no
problem resulting from your voltage reading. I've used power supplies
that don't even have -12V or -5V rails without issue, though some
motherboard's BIOS will show an alert at POST due to it.


Dave
Click to expand...

Thank you all guys, you've help me figure out what to do. I aint gonna do a
thing!
Thanks again for sharing all your knowledge.

Hank
 
Back
Top