PC wont POST after installing loop

[djc5166]

Hello again forums,

Previously I was asking about info on components for getting into watercooling (thread here).

After getting everyting installed, bled, leak tested, and put back into the case, my motherboard will no longer POST. What I'm seeing looks very similar to the boot-loop problems I have read about on my motherboard (Gigabyte EX58-UD3R). The computer starts for 4-5 seconds with all the phase LEDs lit, then shuts off, repeating until I disconnect the power.

Anyone have any ideas to what could be causing this? My guess was the Heatkiller backplate touching the stock 1366 backplate and shorting something out. I haven't disassembled my components yet to see, but would the computer even turn on at all if this was the case? I've also thought about insufficient contact from the cpu block, although I did follow the directions (19.6mm for 200N) for the HK 3.0.

I have tried resetting the CMOS via jumper (as suggested in numerous threads), but still cannot find the problem.

Perhaps the motherboard decided to die coincidently, or could this problem be from the GPUs? Would the mobo still POST with a faulty GPU?

I was really excited about how the loop turned out, but this has turned into a nightmare.

ANY help is appreciated.

 

[henydiah]

unplug Ac power from PSU, remove waterc ooling, remove CPU, clean old thermal paste with alcohol, install again CPU , replace thermal paste , try use Stock cooling, go default system , turn on PC, is that working !

 

[rubix_1011]

Sounds like a ground issue, like OP mentioned. It could also be that there is a BIOS setting that shuts down the system if there isn't a CPU cooler fan detected. Most MB's have a setting that disables this, if this is the case...do you see a message about CPU fan missing/not detected?

I suspect the backplate is shorting against the board, and it may or may not cause it to short right away. You don't have to disassemble the whole loop...you can pull the CPU block and backplate if your case has a backplate access area. Otherwise, you'll have to remove the board.

 

[djc5166]

I have one of my 3 pin rad fans plugged into the cpu fan, would that still trip the feature? Nothing is displayed, it shuts down in 4 seconds.

 

[djc5166]

Took everything apart, ran only the mobo outside of the case with the stock heatsink and RAM. Same bootloop problem...

Is the stock 1366 backplate supposed to have washers on it? I went from a Scythe Mugen 2 to HK 3.0 so I had to put the original backplate back on (needed to be off for the scythe), but don't recall if it came installed with washers. Could this be the problem?

Also, if I leave the extra 8-pin connector disconnected, the mobo stays on. Not sure what this signifies.

Could it also be a power supply issue? Never had, or even heard of anyone having a Corsair HX dying on them...

If none of this works, does this mean the mobo is dead? I was thinking of getting a new one anyways, the UD3R was kind of a budget board IMO. Could the bad board have hurt anything else?

 

[rubix_1011]

If you have a metal backplate, ALWAYS use washers (not metal washers, either)...usually they are nylon/plastic. Some come with a rubber piece built into the backplate that acts as an electrical insulator.

Any time you put metal against your board, you run the high risk of a short. There are thousands of solder contacts that can come into contact with a backplate...and this could cause permanent damage.

 

[djc5166]

That's what I'm asking, if the stock plate had washers for it. It has a rubber face that touches the board, but other than that I don't remember there being washers when I removed it long ago. I assume all the stock plates on all of the 1366 boards are the same.

Obviously I used the washers that came with the HK and my other heatsinks, that's a no brainer. I'm just wondering if the stock one came with rubber AND a set of washers.

 

[rubix_1011]

As long as the metal/backplate didn't make contact, that probably isn't the problem.

 

[apache_lives]

reset cmos
reseat video and ram
try one stick at a time

 

[djc5166]

reset cmos
reseat video and ram
try one stick at a time

No luck...

Also, the system speaker makes 1 faint click sound when the system powers on, and 1 when it powers off.

Edit:

Tested the power supply with a meter, no funny readings, doesn't seem to be the culprit.

 

[rubix_1011]

Did you have any leaks at all? Just wanted to check...water can short just as easily as metal...

Something doesn't make sense...nothing broken off (capacitors, resistors, etc on install) no bent CPU/MB pins...

Do you get anything on the screen before it shuts off? Do you even get to BIOS/POST?

System should still post, but if you have a faulty GPU, it might be dependent upon video and just sit there, but wouldn't necessarily shut down. I had something similar to this for my brother's PC...it was older, though. 1 of the DDR2 channels just went bad, so I had to switch the sticks to slots other than the 1 bad one. Take out all RAM and only try 1 stick in each slot (similar to what apache_lives suggested) unless you already did so. Not just the first slot...but ALL slots.

Might also consider disconnecting all extra hardware...floppy, DVD drive, USB add ons (need keyboard/mouse), 1 video card, no extra PCI cards.

This sounds a lot like a bad ground or a short.

 

[djc5166]

No leaks. I was very happy with how the loop turned out, everything is sealed up tight.

Nothing broken, and no bent pins to my knowledge.

System does not post, and does not get into bios. It just constantly power loops like I have read gigabyte boards to do.

Tried every RAM configuration, nothing else is connected. If it is a faulty gpu, would that of damaged the motherboard? I would not want to buy a new one only to have it be damaged by the gpu.

The same thing happens whether or not the gpu is present, if that means anything. Will a board post with no gpu? I have an 8400gs I can try today when I get home.

Could the motherboard have been damaged simply from the force of putting the tubing on the cpu block? I recall dropping plenty of things onto motherboards that would have had more force than that.

 

[rubix_1011]

You can try the GPU swap...but I would think it would at least still run even if it didn't find a video source. You obviously wouldn't see anything, but I also don't think it would continually boot. If you have more than one PCI-e slot, try that, or try the other GPU to see what you get. If you know it works, likely not GPU.

Do you have your ATX plug seated firmly or all other power correctly seated?

 

[djc5166]

I thought it should post with no video, but then I thought about dead card = no card = no post, but haven't been able to find any info about posting with/without a gpu. I can try the 8400 in my system and the 470s in another. I'm just worried that if a 470 is indeed bad it could destroy the other board too.

Has to be either the motherboard(most likely), gpu(maybe), or the cpu(doubt it).

The connectors are in, I've taken this thing apart at least a half dozen times.

 

[orangejuice789]

My guess would be that in the (dropping of plenty of things onto your motherboard) you broke a circuit? check GPU first though because that is easier to test.

 

[djc5166]

Not that motherboard in particular, I'm just saying I've done things more stressful than putting the tubing on.

 

[henydiah]

Test it in out case .. Assembling again include cpu. Good luck

 

[djc5166]

Test it in out case .. Assembling again include cpu. Good luck

Already did

 

[henydiah]

Test mobo + cpu only + speaker mobo , No beep an no light, try use other cpu , if you don't have , call professional technician pc, I term your psu is fine.

 

[djc5166]

If I did, it was a very long period of time between when it happened and when I powered it up. I had the pc together bleeding for a few days, it would of dried by then I'd think

 

[djc5166]

Wait a minute. You bled on the motherboard?

The waterloop was bleeding...

 

[djc5166]

This is so strange...

I put the 8400GS in -> same bootloop problem

Washers under the stock 1366 plate (the one that is already rubberized) -> same bootloop problem

I put the 470 GPUs in a Asus P5N-D (750i) motherboard (with the 850HX Corsair PSU, Core 2 Duo, and 4GB DDR2 RAM) -> System won't even power on

The mobo has a light that is green when on, so I know it is getting power from the PSU. I know this this 750i mobo, the CPU and the RAM are working, I put them back in the system I took them from and its still running fine (I'm typing this on it right now).

Now it looks like it could still be a PSU issue? The PSU was running (and seemed to test out) fine...

Guess it could still be a GPU issue, but I don't think the PSU in this C2D system is enough to run a 470 to test them.

I'm out of ideas...

 

[rubix_1011]

PSU...has to be. It can't be your CPU or mounting of the block since you get different results with different video cards. You've tested RAM and CPU.

Unplug power cord, shut off the toggle switch on the back. Wait 1-2 mins, plug in, flip the switch back on...see if you get power. Maybe your board isn't reacting well to the hardware swaps.

If you still get the same result, can you test with a different PSU?

 

[djc5166]

I'll try to be clearer, I reread my post and it sounded a bit confusing.

I have another system with a Core 2 Duo, 8400GS, DDR2 ram, and 430w PSU.

PSU...has to be. It can't be your CPU or mounting of the block since you get different results with different video cards. You've tested RAM and CPU.

I've tried both PSUs with my X58 Gigabyte Board, i7 and DDR3 RAM
Both still bootloop

I've tried putting the 8400GS into my X58 board (its pci-e) with the i7 and DDR3 RAM
Still bootloops

No changes with different video cards. The system still bootloops.

I really haven't tested the CPU, as I only have one X58 board.

For the DDR3 RAM, I figure if all 3 sticks went bad, then whatever happened was enough to tank my whole system. I've tried all 3 sticks in different combinations.
Still bootloops.

Mounting the CPU waterblock, the Scythe Mugen 2, or the stock cooler on the i7 (with required washers intact).
Still bootloops

All this is would make sense with the motherboard being bad...

But then I took the 850w Corsair PSU (which tests fine with a volt meter, not the same as real load I know), my 470s, and put them into the C2D system.
The system will not power on, even though the power light is green on the motherboard.

Unplug power cord, shut off the toggle switch on the back. Wait 1-2 mins, plug in, flip the switch back on...see if you get power. Maybe your board isn't reacting well to the hardware swaps.

If you still get the same result, can you test with a different PSU?

Which motherboard? The 750i one?

I already reassembled everything back into my C2D system, I was nervous I might have ruined another board (I built this for my dad) but everything seems to be ok. I can try again tomorrow if you think the system not powering on was just a fluke.

The only other PSU I have is the 430w one in this C2D system. Can that handle a 470? I'm guessing no since it only has one 6-pin pci-e connector.


My friend works at the tech center at the local college here, he said he should have some stuff to test all my components and find out exactly where the problem lies.

 

[djc5166]

I'm sorry but you're still confusing...

Sorry, I'm trying to be as clear as possible.

Your 430w should be able to power your 470s just to load windows or to at lease get into bios to rule out PSU.

Ok, that's what I needed to know. I wasn't sure if the PSU was strong enough even for an idling 470. Like I said I do not want to cause even more damage.

You said you can try each ram stick one at a time and they all work, but when you use them all you can't make it past post? If this is what you mean then it sounds like motherboard still to me.

No combination of any number of my RAM sticks allow my x58 system to get out of the bootlooping.

No combination of any components with my x58 board thus far allow it to get out of the bootlooping.