PSU and mobo woes: A request for help.

[LANjack]

Ok so.....

It all started back in August, when I returned from a trip. Much to my chagrin, I found out that my computer would no longer boot. I tracked it down to either the mobo or the cpu, and went about RMA'ing the mobo back to MSI (Neo 2 plat). October, the board shows back up, and I rebuild the system and the same symptoms happen. The system still doesn't boot. I RMA the mobo again and request a different one, since I had RMA'ed with the "No Power" symptom originally and I figured they hooked up a psu and saw it had power and just sent it back. Soooo.. November rolls around and the board shows up, I let it sit for a while since Im working on another machine. The other machine fails on me (hooray for deskstar hard drives) and I decide to hook the RMA board up and get the AMD system going. So I work for far too many hours setting up the water loop and such, and finally get ready for the fateful moment.. click... click click click... @#$@#$ no boot. But here's the weird parts:

I've tried 3 different PSU's on this motherboard.

A cheapo comes-with-the-case psu, that when plugged in and turned on (the switch on the back) will run and run without ever having pushed the power switch for the computer case. This is a brand new psu, so I don't think it would have any issues.

A thermaltake purepower 480 butterfly psu, that when plugged in and turned on will sort of cycle on and off, it will kick on for a split second and feed power to everything hooked to it (fans, ccfl's, leds, etc) and cut right off again. Click clack, click clack, click clack, infinite.

Another thermaltake , whos model I forget right now, but it is definately the priciest and nicest of the 3. It's usually used in a server, and provides good power. This seems to function correctly in the system. It doesn't turn on until I hit the system power button.

I've hotwired them all, and they all function fine when not connected to anything (including the mobo). So I'm fairly certain all the PSU's are good.

Now my question is.. what the hell is going on?

Is this mobo, fresh from RMA... bad again?
Could it be my CPU causing these boards to fry? (with no chip in, they still exhibit the same symptoms)

I've disconnected EVERYTHING, so I know its not another device thats the problem.

Does anyone know what a constantly running PSU (on without being told to turn on) or the cycling power thermaltake in instance 2, anyone know what that would imply?

Now normally I would suspect the motherboard, but what are the odds that Im experiencing the same symptoms on two completely different boards, both having been RMAd and claimed fixed by MSI.

Anyone have any insight? Im about to throw it all in the trash.. :mad:

[STEvil]

my old neo2 kicked the bucket mysteriously one day....

my guess is the "coldboot" bug is related to a hardware design fault and has finally taken its toll on the board causing it to die completely.

[LANjack]

So should I just cut my losses then?

I just tore the whole thing down... again.. for the billionth time.. and mounted it up in the mobo box on cardboard. Only thing hooked up to it was the atx power connector.. no other hardware inside. I flip the switch on the psu from O to I, and the northbridge fan kicks in.. power supply fans are all on.. and I haven't touched any on/off switch for the motherboard/case. I'm looking at having to do ANOTHER rma.. this will be the 3rd. Thing is, what are the odds that the exact same thing happens with a different board from a couple rma's (different serial #'s, I requested a replacement instead of a repair after the last no fix...). I cant shake the feeling that Im somehow farking these up in the same way each time.. like it's some bad component that Im putting on thats frying it all.

[STEvil]

got a spare cpu to test with?

[LANjack]

Unfortuneately, no. At this point I don't even have a CPU in the board, and it's firing up on its own still. But now that you mention it, it was working when I was testing it out with no components attached waaay back when it came back from RMA the last time. You think it's possible the chip is somehow frying these boards when it gets socketed and juiced? Maybe I need to RMA both at this point...

Oi.. if it weren't christmas time, I would just upgrade to PCI-E, but thats gonna be like a 800$ upgrade (chip, board, gfx, PSU), and cash is always strapped around the holidays.

Oh and I guess the next question is, is there any way to veify if the chip is good or bad without sticking it into another board and risking its life?

And as a side thought.. I've gone over the whole board with careful scrutiny looking to see if theres any sort of hair or smudge that could be bridging traces somewhere, and I didn't see anything on either side.

Could this have anything to do with a grounding issue? I've removed the original HSF bracket in order to mount the TDX hardware, and it was sitting in an acrylic case when it decided it didnt want to work.. or this could simply be coincidence..


Arg.. so frustrating.. got into a beta for a game Ive been dying to play and no system worth playing it on!

[STEvil]

all neo2's will turn on when provided with power, its a design fault..

you need a cpu to get any kind of post screen btw.. no cpu, no brain.

[LANjack]

Well.. I thought it was shorted somewhere.. I thought that's why it would go on without having ever been turned on.. now that you say this should happen? Curious..

I was afraid of burning up a PSU/CPU by leaving it on.. but maybe it's worth another shot..

Thing is, the design flaw doesnt explain is how the thermaltake butterfly is acting all flakey when hooked up to the mobo.. with or without chip, components, whatever.. it sits there and flicks off and on for a split second over and over.. like some sort of short circuit protection is kicking in. I've jumpstarted that psu and it works fine outside of the mobo..

[LANjack]

Still no go..

I put the cpu and gfx card back in.. flip the switch on the PSU and no dice; fans all spin up with alot of enthusiasm as before, but no video output.

The board has to be hosed.. it should post.. give me error codes or something with no CPU in. The question is how, and why 3 times the same problem.

[jjcom]

I would just RMA the CPU and try that...sounds like your CPU is dead to me

[LANjack]

Was hoping you wouldnt say that :(

Looks like I have to rma board and chip then.. maybe I should just upgrade in the meantime and sell off the RMAs when they come.