I have a Compaq Presario SR1910NX.
When I attempt to boot up the monitor's "on" light briefly comes on abefore going off, but I don't get any activity on the monitor's screen. It makes no difference whether or not I use a video card or the embedded video connection on the motherboard.
Can I assume thios means the motherboard is toast and I should just cannibalized the system?
Thanks a lot.
Darren Harris
Staten Island, New Y9ork.
There's a whole checklist to go through,
before you get to cannabalise it.
Is the fan spinning on the power supply ?
Is the CPU cool fan spinning, and the cooler still
securely fastened to the socket tabs ? Sometimes a tab
busts off and the heatsink isn't making good contact.
There have been motherboards in the past, where a flat CMOS
battery (CR2032) freezes things up, even when it should not.
The +5VSB from the power supply, takes the place of the battery,
when the power supply is running, which is why this should not
happen. Yet, some people get operation, when a new battery is installed.
Be careful not to bust off the contact on the battery socket.
http://h10025.www1.hp.com/ewfrf/wc/document?cc=us&lc=en&dlc=en&docname=c00680695
* Manufacturer: Asus
* Motherboard Name: A8N-LA
* HP/Compaq motherboard name: Nagami2L-GL8
* Chipset: GeForce 6150 LE, S939
http://h10025.www1.hp.com/ewfrf/wc/document?cc=us&lc=en&docname=c00683218
It's hard to say if I got the right motherboard. It appears to be an
Asus, perhaps from around the beginning of the capacitor plague. The
board has ordinary electrolytic caps, so you'd check for leaking or
bulging caps. The board would likely have given some warning signs,
such as crashes while you were using the computer. It wouldn't
likely operate all fine and good, and just drop dead. There would be
little hints before that, anomalies that would serve as a warning.
Check that the two power cables are secure. Very occasionally, there
is flaky contact between ATX12V and socket, or between main power cable
and motherboard socket. Make sure both cables are secure and
the latches hold the cables there. You can pull the cable, and visually
examine the shiny pins for a white, less shiny appearance, which means
the pins overheated and the surface metal is damaged.
It might not be getting power. (Check with a multimeter, that
the main power cable is delivering power to the motherboard. The
probe on the multimeter, can slide down into the plastic shell on
the main power connector, and touch the metal pin and take a sample
reading.) I connect the ground of my multimeter, with an alligator
clip, to a screw on the back of the computer, like an I/O connector
screw. That way, I'm only waving one red probe around when taking
voltage samples with my digital multimeter inside.
It might be an onboard regulator which is failed. Checking for leaking
caps, burned MOSFETS (caused by a MOSFET failure), or burned out
toroids (excessive current flow when MOSFET blows), might hint at
damage to the VCore circuit.
*******
With the preliminaries out of the way, turn off all power to the computer
and unplug.
Remove the RAM sticks and place in an antistatic bag. Use a wrist strap
connected between computer chassis and you, to reduce the risk of
damaging the RAM.
When the PC is powered up and you make the next start attempt, the piezo
buzzer thing on the motherboard should beep a RAM error. Because no RAM
is installed. Such a test is used to prove the processor can read BIOS
code, the processor finds no RAM, the processor programs the beeper
to make the resultant sound. If you hear beeps, that is a good sign,
because it means a lot of the motherboard is working. Any test that
generates a beep code would be a good test to use. On one of
my older motherboards, after around two minutes with no keyboard
plugged in, it would beep for that.
Keep an eye on the IDE LED, which flashes for each disk operation,
If no disk access is happening, then that's another sign the BIOS
code is not running, as the POST sequence runs.
Some NVidia chipsets have excessively sensitive PCI Express
slots, and you can blow them out when not using your wrist
strap to reduce ESD risk. Has the video card been changed
lately ?
Is the heatsink still secured to the NVidia Northbridge ?
I see my non-favorite stainless steel mount points, held
in place with the wrong solder. Sometimes that spring wire
unclips from the motherboard, and the heatsink is no longer
secured to the motherboard. The Northbridge could be between
5W and 10W, and having some cooling on it is important.
With no heatsink, it might cook and get ruined.
Some Dells, they have an electrical check for a missing Northbridge
spring clip. They run current flow through the spring clip - if the
clip falls off, the circuit is broken and the computer will no
longer power on. It's a clever way to protect the Northbridge
from cooking because of a spring wire failure. Too bad they couldn't
have found a better way to fit the spring in the first place. They
could have used a tin-lead coated motherboard fastener, so the
soldering would be a bit more secure. You really shouldn't
tug on a solder joint in the first place, as tin-lead solder
is not a structural material. I wouldn't build a house out
of tin-lead solder joints.
*******
So your first real diagnostic test, is a "beep" test. If you get
no beep, it could be a power issue. In which case, I'd use the
multimeter set to 20V DC full scale, to check the main power
cable levels. You can get a pinout for the 24 pin connector,
on page 37 here. Also, check that a standard power supply is in
use, as the wire colors should match the colors listed in the
table. You *can* make measurements on the main cable, while
the plug is in place. The probe barely fits next to where
the wire goes into each "bay" in the plastic connector. As long
as you seek a ground connection away from the action, the possibility
of shorting something with just the red probe, is limited. Multimeters
are available from Harbor Freight for around $20 or so.
http://www.formfactors.org/developer/specs/ATX12V_PSDG_2_2_public_br2.pdf
HTH,
Paul
