dslreports logo
    All Forums Hot Topics Gallery


how-to block ads

Search Topic:
share rss forum feed


Ronkonkoma, NY

CRISIS! Someone please help me determine what is dead!

Ok, going to start with backstory...

Build :

AMD Phenom II 940
Asus M4A79
XFX Radeon 4890
8GB Muskin DDR2-800
Cosair HX620
Noctua gigantic tower heatsink, 2 fans in push/pull
Antec 900-Two Case.
5 Hard Drives, ton of 120mm fans...not important here

I will admit, I occasionally blow my PC out with an air compressor when I don't have canned air. I always watch for it to spit any water...we have several good ones, and they tend to not spray any...but lets not take this out of the equation. I could have simply gotten something wet...but that doesn't really explain much yet. Lets go on..

Lets also add to the equation, that I had to carry this beast up 14 stairs, with a very poor spine, and that it got dropped down 3 stairs. Bent heatsink, 2 broken fans, and the case held together by wires. Yet somehow it was still functional as long as it was kept on it's side, so the heatsink wouldn't begin to lift off the processor...gravity, and the thermal paste kept it stuck and still kept it chilly.

So, my birthday comes, and I decided to rebuild a little bit. What I replaced or added, is as follows.

Case : Corsair Carbide 500
PSU : Corsair AX850
Cooler : Thermaltake Water 2.0 Pro
Also added 2 new Noctua P-12 Fans.

Took a solid day, but I rebuilt the entire system...routed cables to make it look pretty, and it's got quiet, insane airflow like you wouldn't believe. Problem is, when I flipped the power supply on, and pressed the power button, all the fans started going, the PSU fan started to spin and slowed to a stop, the video cards fan stayed at 100%, and 3 of the 4 lights on the top of the video card stayed lit, and no display came on the monitor.

Ok, built dozens on PCs...there are supposed to be issues. So, did a little research...the 3 lights on the video card (RADEON 4890) usually indicate a core overheat, or that the card isn't getting enough power. Tried my old power supply...same thing. Another search turned up that the motherboard may have been touching the case and is causing a short. So, I begin the tear-down. And rebuilt the most simplistic components on a piece of cardboard. Just the motherboard, 1 stick of memory, the CPU, the cooler with the radiator off to the side with 1 fan, and the video card.

I simply cannot isolate the issue here, and figure out which part is causing the issue, or if it's even fixable. I'm very close to ordering a new CPU, Mobo, RAM, and Video Card, which I planned to do for Christmas. I'd have done it all at once, but I just had spinal surgery, and haven't had income for over two months, so I'd rather not drop my savings...so, here are my theories, and what I've done.

I'm wondering which part is somehow now defective...the CPU, Motherboard, or Power Supply.

I have ruled out the video card, since that's the only piece of hardware I had in the house to swap...I used a GTX520, which doesn't require power from the PSU, and same result. And I tried both cards in all four PCI-E slots. Same result.

The Cooler : Installation was simple enough, but the backplate required stick-on pads onto the backplate, connected to the motherboard. I'm not entirely convinced that this is even an issue, since the instructions specify it, but sticking something onto the back of my motherboard makes makes me a little leery. This doesn't seem like the most probable issue, since the motherboard still lights up, and everything seems to work. I don't think this would cause a short.

Power Supply : It's a little late in the day to call Corsair, but the fact that the fan spins then stops, makes me a bit nervous. However, it still powers the motherboard, and spins ALL fans at full speed. And switching power supplies made no difference, aside from my old PSU, that's maybe 6 years old, and was dropped, has a fan that always spins. I'm thinking my new PSU has a fan that ramps up when the temps get higher, but again, too late to call Corsair to confirm. And what has me thinking, is my motherboard has a reset and power button on it. This PSU is 80Plus Gold...they light up instantly. My prior PSU, they would flicker first, then stay lit. Makes me thing this PSU is providing very stable power.

The Case : Motherboard standoffs...it has 6 of the traditional ones, and 3 that are kind of cones that are stamped into the metal. I didn't think much of it, but I have heard of these causing a short in the motherboard on occasion, and you should use washers.

CPU : No clue...something shorted it out? It was never in a position to get hot though...the new cooler made excellent contact, and I just don't how it would cause this issue. And I never smelled that telltale burning smell, nor did I feel any excess heat at any time.

Now we get to my strongest culprit, the motherboard...I have a few concerns, and one may be my fault...I worked late into the night.

First, the 4 or 8 pin secondary power connection on the motherboard. Mine has always had the left 4 pins "blocked" with a little plastic insert, so I always used a 4-pin. New build, I noticed for the first time that it can be taken out, so I used an 8 pin, thinking that maybe I was supposed to remove this in the first place, and use the 8-pin..."Maybe why this is why I could never get such a stable overclock!!" Painkillers, pain, caffeine, and 1AM do not mix. So, theory one, I blew the important things on the motherboard. It will not run the CPU, or send voltage to the PCI-E slots, hence the build not booting, and even the second graphics card not working, since I'm sure the video card does indeed need voltage from the motherboard to work.

Second theory, the drop. Simply put, electronics aren't expected to survive impact. Let alone a 60lb PC tumbling down stairs. I got lucky since I didn't move anything...when I rebuilt and took everything out, maybe something shorted when I did that, maybe something small broke that I'm not seeing, etc.

Last theory...the old heatsink pulled the CPU out of the motherboard socket since the thermal paste was so absolutely stuck to both pieces. Perhaps this damaged a pin, or the socket, causing the motherboard not to post?

I've been working on this all day, getting ready to both pull my hair out, and click the check-out button on NewEgg, spending money I really don't want to. Plus, I'd like to get this build back up, even if it's just temporary, so I can grab a few things from my desktop, AppData folder, and a few other things I don't backup regularly. Who knows what may happen booting up my old drives on new hardware.

Experts, I need your help to narrow down, and isolate if anything is defective. I'm sure I'm leaving out small things I've done to test parts, since this is a somewhat desperate rant, but I've got to get my machine back up somehow. I thank you all in advance for any suggestions or advice.



Indianapolis, IN

Man, you have a hard-on for cooling don't ya? Sorry, I'm not an OC'er.

So....this things works just fine. You take the pieces out, put them in a new case (adding new power supply) and wha-la....it doesn't work? And you put the old power supply in and it still doesn't work.

Is that right?


Ronkonkoma, NY

I do...it's disturbing.

And anyone who's viewed this, I figured it out.

It was the drop that caused the issue, and nothing was actually broken. Here was the issue...

The drop broke the locking mechanism on the socket. I was in the process of giving up, so I started to rip apart everything. Taking the cooler off, something just didn't look right. Turned out, the processor was into the socket...at a slight angle, causing the back portion of the pins to not make contact on the motherboard. But it had no problem "locking". Was a small, yet stupid and major overlook on my part.

On a final note, she's up and running beautifully I might add. I do have an obsession for cooling, and with 5 Noctua P-12 fans (2 front intake, 2 top exhaust, 1 bottom intake), 1 side 200mm intake included with case, and 2 of the nice Thermaltake fans in a push/pull on the radiator on the rear exhaust slot, we're running at 35* C idle, 42* full load, after a 10 hour Prime 95 Torture Test on the CPU/FPU only...this thing is rock solid stable, and chilly. Now, it's time to overclock


Indianapolis, IN
reply to staticwinter

Troubleshooting can be so slow and tedious sometimes.

Glad you got it.