PC won't POST - anything left to try?

fwor

I'm trying to fix a PC for someone, and I've run out of things to try.

It "just died" (i.e. it didn't stop working as the result of something being changed).

It will power-up, and will run up the fans and HDD, but that's about it. No beep codes.

I've removed all PCI cards (the screen is now connected to the built-in graphics chip), tried removing either of the two memory sticks, and tried each stick on its own in both DIMM slots. Disconnected HDD, CD-ROM and floppy. Checked BIOS battery (good) and shorted the CLR-CMOS link to reset BIOS to defaults.

None of these have made any difference.

I've checked the PSU and all voltage rails are within ATX spec - the PWR-GOOD line is high (+5V).

Any suggestions? I'm loath to give up on it, as the owner can't afford a new one...

dogmaryxx

Not my area but after some googling the only thing to try seems to be

take out the RAM and attempt to power on the PC (you can't expect it to fully boot without RAM, but we just wanna test something here...), if you get beeps or a "no RAM error", then you have your answ: Your RAM has been messing with your PC, and you may need to buy new RAM. If you have more than one RAM stick and are feeling bold, you may try to figure out which one was the one giving you problems via trial and error (but keep in mind first that it is quite possible for both RAM sticks to be dead)

fwor

Thanks - just tried that, but no change and no beep codes, so it doesn't seem to be getting as far in POST as memory checking...

spakkker

Not getting any beeps does point to m/board. However I think the psu is always number 1 suspect , can you swap this ,in or out?
Trying the m/board outside of the case is often the next step.

Keep an eye out on ebay and local classifieds for something similar , esp. with hdd removed but which posts.

Mista_C

Another score for motherboard here. It was/is a common problem with the Intel 965 boards iirc.

fwor

If it were a new build or if it had been fiddled with I'd consider a short to the casing, but the probability of that happening out of the blue seems to be very low.

Sadly I don't have a spare PSU of the right type to swap in - hence the careful measuring of rail voltages. But I agree that voltages alone are not a complete enough test - if there was excessive ripple for example then I doubt it would run.

It's a Winfast mobo, and there are a couple of the same type on ebay, but I can't bring myself to fork out £40+ for a secondhand 6 year old motherboard...