System not working

[x88x]

Before I tear down my WC loop again, I thought I'd throw this out there and see if anyone had any recommendations.

I finally got my system back together, after ~4 months, but it's not working. It does power on, fans power up, lights come on, but nothing shows on the monitor. I dug up a system speaker that I know works and hooked it up to the speaker output, and it isn't giving any diagnostic beeps.

For some context, this is the recent history of this system:

##Last time this system was working##
Slow coolant leak killed sound card and MBB.
MBB replaced.
2nd GTX260 added.
##System reassembled##

The fact that it is powering on at all tells me that the PSU and CPU are good. I have tried it with both 260's plugged in, each 260 individually, and a GTS250 that I happened to have on hand (known good). I have pulled out the RAM in chunks of 2, and am shortly going to be shutting down my server and pulling the known good RAM from that to test. I have pulled the CMOS battery to do a complete wipe of the BIOS. Oh, and I have disconnected the SSDs, just on the off chance that they were affecting something. Nothing so far has changed the behavior.

At this point, I'm afraid it's the MBB (known good RAM from the server will be the last straw), but I wanted to see if anyone else had any ideas, or noticed something I didn't.

[x88x]

Checked with the known good RAM. Same behavior. :(

Any ideas? At this point I'm fairly well convinced that it's the MBB. Last thing would be to drop the CPU into my server (only other S775 system) to make sure it still works.

[msmrx57]

Well, I'd either pop the CPU into the server to check it, or use the CPU from the server to check the board. Or both. Also did you double check all the connections from the PSU? Other than that I think you pretty much covered all the bases.

[x88x]

I checked the power cables and they're all good. I'll be swapping the CPU into my server (suspect part in a know good system instead of vice versa) tomorrow...gonna get to bed now, see if I can actually get to work at a reasonable hour tomorrow. :P

[Konrad]

You said that you pulled the mobo battery to clear CMOS ... but do you know if the battery is still good?

Does the mobo have any fuses which might have blown?

Does the system speaker issue a boot beep?

Checked all your jumpers?

[x88x]

You said that you pulled the mobo battery to clear CMOS ... but do you know if the battery is still good?
I'm reading 2.9V...it should be fine there, I would think.


Does the mobo have any fuses which might have blown?
Nope.


Does the system speaker issue a boot beep?
Nope.


Checked all your jumpers?
Have now..they're good. I even tried it with the CMOS clear jumper in both settings.

I tried my CPU in my server and (thankfully) it worked just fine! :D At least that means the single most expensive component is fine.

Unless anyone has another idea, I think the MBB is toast. ;(

[Konrad]

No boot beep at all is bad. You could try measuring volts at various test points, try to trace where the circuit goes bad ... it might be a dead electrolytic, bad connector, or some other trivial component. Or a piece of zorched silicon, impossible to repair. I'd say if you can visually spot (or identify through touch/temp or smell) and easily replace the cooked part then go for it, otherwise it's not worth the time/expense.

Too late to RMA?

[x88x]

Actually, I bought it from a forum member. I sent him a PM to determine a course of action.

I haven't tried measuring voltages, but I did a visual inspection of the board and can't see anything wrong with it.. I swear, I've had the worst luck with MBBs with this system...

[Konrad]

What brand/model MBB? There might be a recall or internet chatter about common failure points. Perhaps you need to "emergency" restore bad firmware, your MBB might even have a backup BIOS chip. My experience is that MBB hardware failure is almost always:
- bad caps or regulators; easily repaired at the component level once isolated through troubleshooting
- blown fuses; usually near the ATX connectors, regulator ICs, or VRM stuff; these days they are usually hardwired semiconductor devices which require a little SMT rework
- broken traces, leads, or solder joints, often damaged in transport, usually repairable (if you don't mind ugly)
- ground fault; something is shorting power where it shouldn't, might be repairable
- catastrophically overheated large ICs (often the NB/MCH, especially GMCHs); might be repairable with better cooling, might be permanently cooked (ie: dead MBB unless socketed)
- crystal oscillators sometimes fail inexplicably; they can be tested OOC and are easily replaced

I'm assuming you've already swapped CPU, PSU, RAM, and other off-board components to isolate the problem to the MBB? It never hurts to replace the CR2032 battery (about $1), strip the system down to minimum config, and try a different PSU. There's always the chance of some oddball compatibility problem/quirk with any particular hardware card or drive, computing hardware is never 100% perfect.

[billygoat333]

mmm... brains! /thread necromancy.

[x88x]

Hell, I forgot I even made this thread. :P If anyone didn't catch it in other threads long ago, the MBB was confirmed dead, so after looking around trying to find a dual PCIe x16 board (and failing miserably), I tossed it, sold the CPU and RAM, and built a 1090T system.