low power from 12v rail

A

atblock

Hi, i've had this problem for a while and it's survived through two
power supplies....
My first power supply I had for a year and the only problems my
computer was having was that it would freeze every once in a while. I
don't know what caused this problem, and I never bothered to check the
power supply. Recently, I reformatted and reinstalled windows and all
of a sudden my computer started going into standby at random times
(usually when there was a lot of HD activity or CPU activity, but not
always). My CPU temp peaked at about 50c. My power supply was only
giving 11.8~ amps on the 12v rail. I replaced the PSU, and same thing.
Then I replaced it again with a 400W PSU, and same thing. My 12v rail
won't give more than 11a even though it's rated at 20a....
Motherboard: ABIT KV7
CPU: Sempron 2500 1.77 GHz
RAM: 512 DDR400 Mushkin
Video Card: RAGE 128 PRO Ultra GL
Power Supply: Athenatech 400W
3.3V:28A
5V: 40A
12V1: 20A
12V2: 20A
-12V: 1A
-5V: 0.8A

CPU Temp is 54C
SYS Temp is 56C
AUX Temp is 42C

CPU Fan is running at 3309

Voltages:
VCoreA 1.6
3.3 3.31
5 5.08
12 11.9
-5 .63
-12 -7.01
5VSB 4.85
VBAT 3.23

When I run off of a live linux OS (running only off of a CD everything
seems to be fine)...my problems only happen when running XP
 
U

UCLAN

My first power supply I had for a year and the only problems my
computer was having was that it would freeze every once in a while. I
don't know what caused this problem, and I never bothered to check the
power supply. Recently, I reformatted and reinstalled windows and all
of a sudden my computer started going into standby at random times
(usually when there was a lot of HD activity or CPU activity, but not
always). My CPU temp peaked at about 50c. My power supply was only
giving 11.8~ amps on the 12v rail. I replaced the PSU, and same thing.
Then I replaced it again with a 400W PSU, and same thing. My 12v rail
won't give more than 11a even though it's rated at 20a....

Methinks you are confused. How were you measuring 12v *current*? If you
were measuring "11.8" on the 12v rail, chances are you were measuring
11.8 VOLTS, which is in ATX specification and would not cause your
problem. Measuring PSU DC current is a difficult task. You were probably
using a DVM, in which you were measuring 11.8 volts, not 11.8 amps.
 
K

kony

Hi, i've had this problem for a while and it's survived through two
power supplies....
My first power supply I had for a year and the only problems my
computer was having was that it would freeze every once in a while. I
don't know what caused this problem, and I never bothered to check the
power supply. Recently, I reformatted and reinstalled windows and all
of a sudden my computer started going into standby at random times
(usually when there was a lot of HD activity or CPU activity, but not
always). My CPU temp peaked at about 50c.

Are you certain it is a reading of 50C for CPU, not
something else? I only ask because some generic software
(like MBM5) may need configured to show the right temp for
the right part(s).

You might check the Windows Event Viewer.

My power supply was only
giving 11.8~ amps on the 12v rail. I replaced the PSU, and same thing.
Then I replaced it again with a 400W PSU, and same thing. My 12v rail
won't give more than 11a even though it's rated at 20a....


How did you meaure the 12V rail?
It should read close to 12.0V, ideally, with a multimeter.
If using bios or other software, a reading that is a little
low like 11.8 is not necessarily a sign the power is low,
only that the voltage dropped 0.2V between the main power
rail and the small track going to the voltage sensor. In
other words, on a typical motherboard with a PSU outputting
12.0V at the connector (the ideal), it will read slightly
lower than 12.0V. It is not a problem.

However, if your PSU were generic junk, you might not be
able to trust the amperage, or the voltage fluctuations
under load. Many don't realize the impact of dynamic load o
on a PSU.

You didn't mention your PSU makes and models though so we
are lacking info there. Even so, the initial impression is
that it isn't a problem.

You should do a basic inspection of the system, checking
that cards and cables are well-seated, no capacitors
failing, fans all work and not too much dust buildup, etc.

Had the system first worked ok and if so, what had changed
prior to the onset of the problem?

Motherboard: ABIT KV7
CPU: Sempron 2500 1.77 GHz
RAM: 512 DDR400 Mushkin
Video Card: RAGE 128 PRO Ultra GL
Power Supply: Athenatech 400W
3.3V:28A
5V: 40A
12V1: 20A
12V2: 20A
-12V: 1A
-5V: 0.8A


Oops, I should have read ahead. I don't know for certain
but expect the Athenatech PSU isn't so great. It appears to
be a low-end generic that might be worth closer to 250 real
watts. That would make the above labeled PSU specs either
momentary peak values (which I consider fraudulent) or are
outright fiction. It's not able to output 40A of 12V, we
can be fairly confident of this, but your system is not very
power hungry either. So it's a big variable, I can't
outright blame it as it "might" be barely able to supply
enough power but if I had a better PSU handy I'd certainly
try it (but I still don't know what other PSU you had
tried).



CPU Temp is 54C
SYS Temp is 56C
AUX Temp is 42C

System temp seems a little high, you might touch-test your
southbridge to see how hot it is, consider putting a
heatsink on it if it's bare.

CPU Fan is running at 3309

Voltages:
VCoreA 1.6
3.3 3.31
5 5.08
12 11.9
-5 .63
-12 -7.01
5VSB 4.85
VBAT 3.23

When I run off of a live linux OS (running only off of a CD everything
seems to be fine)...my problems only happen when running XP



Try turning off power management in windows and the bios.
Boot to safe mode and see if it still happens. Does it
_ONLY_ happen if aggressively using the system or also when
it's sitting idle?

is your video card passively cooled? If so, or regardless,
take off the side panel and point a desk fan at the
interior. Is your room ambient temp higher now than other
times of year and/or do you think there might be a
correlation between room ambient, the load-induced heat in
the system, and the crashes?

Also be sure to check the PSU exhaust fan. Often on the
cheaper PSU they need lubed, else eventually will fail to
spin up when system is turned on which is the worst failure
because they're quiet, just baking away until they pop.
 
P

Paul

Hi, i've had this problem for a while and it's survived through two
power supplies....
My first power supply I had for a year and the only problems my
computer was having was that it would freeze every once in a while. I
don't know what caused this problem, and I never bothered to check the
power supply. Recently, I reformatted and reinstalled windows and all
of a sudden my computer started going into standby at random times
(usually when there was a lot of HD activity or CPU activity, but not
always). My CPU temp peaked at about 50c. My power supply was only
giving 11.8~ amps on the 12v rail. I replaced the PSU, and same thing.
Then I replaced it again with a 400W PSU, and same thing. My 12v rail
won't give more than 11a even though it's rated at 20a....
Motherboard: ABIT KV7
CPU: Sempron 2500 1.77 GHz
RAM: 512 DDR400 Mushkin
Video Card: RAGE 128 PRO Ultra GL
Power Supply: Athenatech 400W
3.3V:28A
5V: 40A
12V1: 20A
12V2: 20A
-12V: 1A
-5V: 0.8A

CPU Temp is 54C
SYS Temp is 56C
AUX Temp is 42C

CPU Fan is running at 3309

Voltages:
VCoreA 1.6
3.3 3.31
5 5.08
12 11.9
-5 .63
-12 -7.01
5VSB 4.85
VBAT 3.23

When I run off of a live linux OS (running only off of a CD everything
seems to be fine)...my problems only happen when running XP

Where ever those SYS and AUX temperatures are being measured,
they are quite high. Either those readings are not correct, or
your system does not have a proper cooling solution for the
computer case. In addition to the fan on the back of the
power supply, you should have at least one other exhaust fan,
to remove the hot air from the computer case. An intake fan
placed on the front of the computer case can help too, but not
all computer cases have picked a good location for one.

The quickest way to help the cooling on the computer case, would be
to remove the side cover, and see what happens to the measured
temperatures. If they drop down to say 30C-35C, that tells you
your case cooling was pretty bad. One way to partially fix it, is
to pop out a disk drive plastic front cover, leaving a hole
for cool air to enter the computer. In that way, if you have
an exhaust fan on the back of the computer, it doesn't have to
work quite as hard.

On the power supply, the "volts" are supposed to be a constant.
A 12V output should not vary by more than 5% or so. That means
valid values range from 11.4V to 12.6V.

As the computer works harder, it draws more current. Your 12V
output can supply current from 0 amps all the way up to 20 amps.
There is no measurement device on the motherboard to measure
current, so there is no way to determine how many amps are
flowing. When building computers, we estimate the consumption
of various pieces of hardware, using measurements made by
web sites like Xbitlabs or other sources. That is the only
way to estimate whether a power supply will be adequate for
the job or not. If you were to draw more than 20A, what you
would observe, is the voltage level (something you can measure)
would drop below 11.4V as the current flow goes over the maximum
allowed. The power supply would eventually shut off, if
tortured enough.

For testing computers, two programs you can use are:

1) Memtest86+ , a memory tester from www.memtest.org
This prepares a blank floppy (or a CD version if you don't
have a floppy), and you boot with the floppy once the
memtest86+ installer has formatted the floppy and put
the test program on there. Two full passes error free
is a good enough test. If you see a whole bunch of errors
scrolling down the center of the screen, then your memory
is having problems for some reason.

2) Prime95 , a test program from mersenne.org . This is a
Windows program, and it has an option called the Torture
Test. It carries out a calculation with a known answer, and
if the answer deviates, an error message is printed on the
screen. A properly tuned computer will run this error free
for hours. Prime95 tests processor, memory, and the
Northbridge that connects them. It can identify a problem
with the memory, and if it errors out in only seconds, that
is telling you that your processor is no where near being
stable. That happens to me, if I overclock my mobile AthlonXP
above 2400MHz :)

Try those two tests and tell us what the results are.

If you have one stick of memory, and that motherboard has
three memory slots, the memory stick should go in the slot
furthest from the processor. That puts the memory stick
right next to the bus terminators.

I believe your processor is a Model 8 Sempron, FSB333 and
core of 1750MHz. The multiplier would be 10.5x. The best
setting for the RAM would be a choice of DDR333, not
DDR400. By selecting DDR333, the memory stick and the
processor run synchronous, and by running synchronous the
Northbridge can typically use a lower latency communications
path. You will have improved memory stability and better
overall performance. At least, that is how I would set it
up.

You can check your vital statistics first, by using CPUZ.
This will tell you what clock rates are being used, and then
you can decide whether you need to enter the BIOS and
adjust stuff.

http://www.cpuid.com/cpuz.php

http://www.cpuid.com/download/cpu-z-135.zip (the download)

HTH,
Paul
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top