no power after replacing psu: why?

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

Threaded View
Hi, I recently replaced my cmos battery and the power supply for a
relatively new machine.

I heard no post and windows didn't load, but after I replaced the
battery, Windows loaded normally.

Next I replaced the power supply unit. Let me say I don't have a lot of
experience with that, although I am relatively careful when poking
inside my PC. The first time the power went on briefly, but stopped,
leading me to believe that the cpu fan power is not working or
something like that.

I am not terribly worried about the  motherboard  because it sounds as
if something is not connected properly or  touching something  that's
not supposed to.  but I can't figure it out!

How do you troubleshoot these issues?
(here's an online resource: )

I've removed all connectors to the hard drives and disk drives, and
only connected: mobo atx power connector, cpu fan to mobo CPU_FAN and
PWR_FAN and chassis fan. I have a power light that goes on when
connected, except that nothing turns on.

1. what is a multimeter, and why/how would that help to identify this
kind of problem? Can you still do it without one?
2. I'm using brand new cables, and the only wierd thing about the psu
is that it was a 24 pin atx power connector for a 20 pin atx mobo. The
psu instructions told me how to separate 2 rows (4 pins) which I did
easily enough.

Either I'm missing something obvious or one of the cables are bad. Any
other possible things to look out for?


(for more detail see
further description of cmos/ posting problems

here's my specs in case you're interested:
AMD Athlon 64 3500+        CPU-ADA3500AWBOX            1   286.00
2.2GHz Microprocessor,
Socket 939, L2 512K,
ADA3500AWBOX. Retail Box
Asus A8V Deluxe Socket     MB-A8V-Deluxe               1   132.00
939 ATX Motherboard,
AGP8x, 1394 & USB2.0, 8ch
Audio & GB LAN, S /
OCZ 1GB (512MB X 2)        DIM-OCZ4001024PDC-K         1   183.99
PC3200 400MHz Premier
Dual Channel
Xion Mid-Tower ATX Case    CS-Xion                     1    69.99
*Choose Color = Blue
w/ Side Window, LED
Lights & 450W PS, Color:
Blue, Silver, Black
Zalman CPU Cooler,         COOL-CNPS7700-CU            1    42.99
CNPS7700-Cu, Pure Copper,
Support Socket Intel
Pentium4 775/ 478, AMD
754/ 939/ 940. Retail box
Asus V9520Magic/T/128      VD-V9520Magic-T128          1    65.99
Video Card, GeForce
FX5200, 128MB DDR, AGP8X,
TV-Out, Retail box.
Western Digital Caviar SE  HD-WD1600JD                 2   102.00
Serial ATA 160GB Hard
Drives (WD1600JD), 8MB,
7,200RPM, OEM

Plextor SATA PX-712SA DVD  DVR-PX-712SA                1   114.99
Dual Burner, 8MB Buffers,

Re: no power after replacing psu: why?

Quoted text here. Click to load it
Quoted text here. Click to load it

When you put the original power supply back, what happens ?

What are the odds that the new supply is DOA (dead on arrival) ?

When you push the Power button on the front of the
case (or however you have it rigged while testing),
do the fans spin ? That tells you +12V is present,
and suggests the power supply has started.

I would give a procedure for testing the power supply, but other
posters don't recommend starting a power supply without a load
on it. To eliminate some other failure, I'd just swap the old
supply back in and make sure things still work (in some fashion).

A multimeter can be used to measure the voltages coming
from the power supply. If you happen to own one, it might
save you having to swap something when it isn't needed. You
can get by without it. I actually test power supplies before
I use them, using a dummy load circuit and a multimeter, and
that is to help eliminate the risk that a crazy PSU burns
my shiny new motherboard :-)

While you might have no interest in testing power supplies,
I did post an article on how to build a load box for one.
This kind of article is suitable for people who like to tinker.


Re: no power after replacing psu: why?

idiotprogrammer wrote:

Quoted text here. Click to load it

Have you tried removing the mobo and PSU from the case and trying them
outside?  Put the mobo on a surface that doesn't conduct electricity
(no metal) or generate static electricity (no plastic, glass, or shiney
paper; formica, newspaper, or wood are OK).  You'll have to prop it up
about 1/2" to prevent the graphics card from popping up.  Instead of
the on/off button, you can momentarily touch the two power-on header
pins to turn on the mobo.  If the mobo works outside the case but not
inside it, you can be pretty sure that you have a short between them.
Check each mounting hole on top and bottom, and use a fiber insulating
washer wherever a hole isn't surrounded by a copper donut ring or has
at least 1/4" of free space all around.  Support the mobo at every
hole, even those that don't match up with any in the case.  IOW for
those you can simply insert a standoff (brass or plastic -- plastic may
have to be cut to the right length) into the mobo (insulate it with
fiber washers as needed), and let it sit against the case.  If you
can't fit a standoff near a corner, stick a bumper foot against the
case there to prevent shorts if the case flexes too much..

Quoted text here. Click to load it

That looks useful.

Quoted text here. Click to load it

Probably the best $3-25 you'll ever spend for a diagnostic tool.  Just
be sure that it doesn't use wierd batteries (some need wristwatch
cells) and that it has at least 3.5 digits of resolution (at least 2000
count) because 2.5-digit meters (200 count) can be inaccurate at
certain readings.

Quoted text here. Click to load it

A meter can tell you if a PSU is bad, but it may not be that useful for
checking for shorts between the mobo and case.

Useful FAQs:

     How to use a multimeter:

     How to check your PSU:

You can turn on the PSU without a mobo by connecting its big
connector's green wire to any of its black wires (there's one on each
side of the green) with a bent paperclip.  If you're worried about
causing damage this way, buy a small resistor from Radio Shack,
something rated anywhere from 100-500 ohms, and use it in place of the
paperclip.  If the PSU fan doesn't spin, then maybe the PSU is designed
to need a load on the +5V rail, which you can usually provide with a
hard drive.   Otherwise try a 5-10 ohm resistor rated for at least 5
watts (10W is better -- resistor won't burn or melt stuff), and connect
it between any red wire and any black wire).

Did you plug the square ATX12V connector with the two yellow wires and
two black wires into the mobo (not the connector  that splits off from
the big mobo socket, which has one each black, yellow, orange, and red
wires)?  Almost any mobo that requires the former won't power the CPU
without it.

Quoted text here. Click to load it

I'm not familiar with Xion PSUs, but the pictures at NewEgg show no UL
registration numbers, which is never a good sign for quality, and many
cheapo PSUs fall far short of their ratings. .  But your equipment
doesn't require too much power, probably < 150W

Re: no power after replacing psu: why?

You don't know what a multimeter is and you're screwing around inside
your computer?  Do yourself a favor and take it to a technician before
you hurt yourself.

idiotprogrammer wrote:
Quoted text here. Click to load it
Quoted text here. Click to load it

Re: no power after replacing psu: why?

Newfdog wrote:

Quoted text here. Click to load it

I wouldn't trust many computer "technicians" to know how to use a
multimeter either.  He can probably replace all the bad parts himself
for less than what a computer shop will charge if just one of them is

Re: no power after replacing psu: why?

idiotprogrammer wrote:
Quoted text here. Click to load it
Quoted text here. Click to load it
If the system loaded in a normal fashion after replacing the battery,
why did you r&r the power supply?


Re: no power after replacing psu: why?

To answer, Logan,

Why did I replace the PSU? the psu was making wierd whirring noises
which were driving me crazy!  I had some suspicion that the PSU might
have something to do with the CMOS depleting so quickly.

After giving up, i brought the PC to the shop, where the repairman hit
a single button and caused it to work. (Luckily, he didn't charge me

When testing the system after replacing the psu,  I was using the
switch on the back instead of the switch in the front.

The switch in the back (the psu switch) didn't actually turn the
machine on, but simply supplied power. I had no idea there was a

Yes, you can start laughing at my idiocy now!  (Actually, this is a
good example of how something which  seems obvious is not as obvious as
you would think).

robert Nagle
Houston, Texas

Site Timeline