[RHTrix]

Hey everyone, I have recently installed a new motherboard, CPU, CPU, and 2 sticks of 4GB RAM.

When I hit the on button, the lights and the two fans on my case turn on for less than a second and the whole computer just shuts off. The CPU fan does not move at all during this short time frame. The whole thing just won't turn on.

So this is what is new in my setup:

CPU: i5-2500K LGA1155
GPU: NVIDIA GTX460
RAM: 2X4GB CORSAIR VENGEANCE RAM
MOTHERBOARD: GIGABYTE GA Z68MA-D2H-B3

This is what I already had on my old setup that I am not replacing:

PSU: TX650W CORSAIR

I also have a hard drive that I have never touched since receiving this computer two years ago.

So I installed the new motherboard, connected all the cables from the PSU onto it. CPU fan is installed and also connected onto the motherboard via cable. I am almost 100% sure I have the F_PANEL cables connected correctly.

I just can't figure this out. The people who have tried to help me are telling me that the mobo might have been DOA or the PSU is messed up, but I am positive the PSU is fine.

Thanks a lot in advance guys!

Edited by RHTrix, 07 October 2011 - 10:17 PM.

[Advertisements]

the first thing to suspect is the mobo standoffs are incorrectly placed or there is one from the old mobo still installed and touching the new mobo. Check this out and if all is OK proceed with the below.

Look up the beep codes for your mobo if they are not in the manual and write them down.

Turn off power but leave your power cable plugged in (it earths your case through the cable earth).

1) Remove ALL your RAM.

2) Remove your Video Card

3) Disconnect your screen, keyboard and mouse.

4) Leave only your CPU installed.

5) Turn your computer on and wait a bit (5 to 10 sec.).

Are you getting a "RAM problem" beep code?
Are you getting "CPU problem" beep code?
If the CPU is OK you should get a "RAM problem" code.
If you get nothing, you have a dead mobo, but that is assuming your PSU is OK. Do you know that as a definite? There is a quick and dirty way to test the PSU, but it would be totally inappropriate for me to tell you how, so do you have a spare PSU you can use?


Make sure your turning off your power switch before each of the below steps and do not do this on a carpeted floor. Keep yourself grounded with your forearm on the case.

If you get a RAM problem code, install 1 stick of RAM in the No. 1 slot.
If the ram stick is OK you should now get a "Video Card problem" beep code.
If so, install your next stick of RAM n No.3 slot.
If the RAM is all OK, you should get the Video card problem again.
Install the Video card. If the video card is OK you should get 1 beep only (post beep).

That's the way it should happen with a new build. Testing one component at a time saves a lot of grief and troubleshooting.

It does sound like a dead mobo though if you do not get a CPU problem beep code when it is the only thing connected. If you get a CPU problem code, then there is your culprit.

Edited by Akabilk, 08 October 2011 - 01:19 AM.

[Akabilk]

the first thing to suspect is the mobo standoffs are incorrectly placed or there is one from the old mobo still installed and touching the new mobo. Check this out and if all is OK proceed with the below.

Look up the beep codes for your mobo if they are not in the manual and write them down.

Turn off power but leave your power cable plugged in (it earths your case through the cable earth).

1) Remove ALL your RAM.

2) Remove your Video Card

3) Disconnect your screen, keyboard and mouse.

4) Leave only your CPU installed.

5) Turn your computer on and wait a bit (5 to 10 sec.).

Are you getting a "RAM problem" beep code?
Are you getting "CPU problem" beep code?
If the CPU is OK you should get a "RAM problem" code.
If you get nothing, you have a dead mobo, but that is assuming your PSU is OK. Do you know that as a definite? There is a quick and dirty way to test the PSU, but it would be totally inappropriate for me to tell you how, so do you have a spare PSU you can use?


Make sure your turning off your power switch before each of the below steps and do not do this on a carpeted floor. Keep yourself grounded with your forearm on the case.

If you get a RAM problem code, install 1 stick of RAM in the No. 1 slot.
If the ram stick is OK you should now get a "Video Card problem" beep code.
If so, install your next stick of RAM n No.3 slot.
If the RAM is all OK, you should get the Video card problem again.
Install the Video card. If the video card is OK you should get 1 beep only (post beep).

That's the way it should happen with a new build. Testing one component at a time saves a lot of grief and troubleshooting.

It does sound like a dead mobo though if you do not get a CPU problem beep code when it is the only thing connected. If you get a CPU problem code, then there is your culprit.

Edited by Akabilk, 08 October 2011 - 01:19 AM.

[phillpower2]

Hi RHTrix and
Before you perform the testing suggested by Akabilk please confirm that the CPU fan is connected to the header on the MB and that both the 24 pin main power connection is securely in place, the same check should also be done with the 4 pin ATX connection that powers the CPU.
Did you only use the correct amount of TIM? See the tutorial provided courtesy of Digerati;
http://www.geekstogo...rface-material/

[RHTrix]

Thanks for the responses guys.

I checked with many people about the thermal paste and they say that a vertical line is best for my processor so I did that instead of making a pea-shape with the paste.

Also, I performed the CPU test. I took out my RAM and my GPU, but the same thing happened. My computer turns on for less than half a second and it just dies out. No beeps, nothing. I don't have a spare power supply either :/

[phillpower2]

Hi RHTrix
What about checking the power connections and the CPU fan connection?
Most modern MBs have a thermal sensor on them to protect the CPU, an incorrect amount of Tim or a disconnected CPU fan will cause the CPU to heat up almost instantly and the sensor shuts the system down just as quickly to avoid frying the CPU.
If the above checks out the three main suspects are the MB, the PSU or less likely unless damaged when installing the CPU.

[RHTrix]

Hi RHTrix
What about checking the power connections and the CPU fan connection?
Most modern MBs have a thermal sensor on them to protect the CPU, an incorrect amount of Tim or a disconnected CPU fan will cause the CPU to heat up almost instantly and the sensor shuts the system down just as quickly to avoid frying the CPU.
If the above checks out the three main suspects are the MB, the PSU or less likely unless damaged when installing the CPU.

Yeah the power connections and the CPU fan connection is definitely fine.

All signs point to the mobo or PSU being dead, but I don't think my PSU is affected at all. It was powering my old setup for a year just fine. I guess I'll just have to send in this mobo as DOA and get a replacement and from there I'll make my next decision unless you guys have more advice or suggestions.

[Akabilk]

For testing purposes, a correctly installed CPU and connected HSF will work fine even without thermal past if the mobo and CPU are OK.

Thermal past is designed to fill "microscopic" pits between the CPU ad HSF surfaces to provide the best possible contact, not some sort of electrical contact pad. Without it there is still plenty of contact for a short CPU test and even if left on for an hour, the CPU temps will be higher than normal, but never get high enough to shut-down the mobo as long as it is not under load conditions.

If the mobo will not run with a correctly mounted CPU and correctly installed HSF, either the mobo, CPU or PSU is dead or faulty.

If the PSU was working with the old mobo and CPU, there is no reason why it should be suspect now as this is only an upgrade not a completely new build. PSU's don't "die" while not connected to anything.

Hey everyone, I have recently installed a new motherboard, CPU, CPU, and 2 sticks of 4GB RAM.

Mobo or CPU?

I'd test the CPU in another (known to be working) board with the correct socket or have it tested. If it works you have a faulty mobo, which is far more common than a faulty CPU.

Just my 2c worth.

Edited by Akabilk, 09 October 2011 - 02:08 PM.

[phillpower2]

Without it there is still plenty of contact for a short CPU test and even if left on for an hour, the CPU temps will be higher than normal, but never get high enough to shut-down the mobo as long as it is not under load conditions.

The OP has not asked for a debate on the pros and cons of the properties of TIM but the above statement needs to be clarified.
A CPU should never be used without having Tim applied and as I said previously most modern MBs have a thermal sensor that will shut the system down to prevent the CPU from frying meaning that the processor will not be run without TIM for anything more than seconds in any event, the only way to alter the sensor settings is in the BIOS and why would anyone wish to do that.
Options before returning the MB as DOA is having a Tech check your present PSU and remove the CPU, check it for bent pins, apply the correct amount of TIM, reassemble a barebones set-up outside of the case on a piece of cardboard larger than the MB, only attach the PSU, the video, 1 stick of Ram and the keyboard, let us know the results.
Dependant on your MB you may need to short out the two power on pins that connect to the switch on the front of your case.

[Akabilk]

A CPU should never be used without having Tim applied and as I said previously most modern MBs have a thermal sensor that will shut the system down to prevent the CPU from frying meaning that the processor will not be run without TIM for anything more than seconds in any event, the only way to alter the sensor settings is in the BIOS and why would anyone wish to do that.

I might suggest that is an erroneous "belief" not a fact. It might have been true back in the day before thermal sensors but it is not now (try it and find out).

Every CPU I put into a new mobo I test without thermal paste just in case the mobo is faulty and I have to clean it all off. If it doesn't work I put in a spare CPU I know works, but every time it has been the mobo.

And I did say "For testing purposes..." (not normal use). A mobo will only shut down above a certain temp. detected by the thermal sensor (there is no such thing as a thermal paste "sensor" only a temp. sensor).
That temp. will never be reached during a brief test and it's impossible to fry a CPU anyway with a modern mobo, it would just shut down. A CPU without paste being tested does not reach a shut-down temp. There just isn't the load.

Dead new mobo's are tediously common and Arctic Silver 5 is expensive (I build and troubleshoot a lot of PC's). Just a few days ago I (finally) got around to upgrading my own board and RAM to DDR3 and a new CPU. I test the new components outside the case before installing including testing the CPU without thermal paste (it has it now though) and that is the computer and CPU I'm using right now.

My 2c worth.

[Digerati]

I disagree with this suggestion to power up a motherboard and CPU without TIM between the mating surfaces. Even for short testing purposes, the risks are too great. There are just too many assumptions that should not be made.

• It assumes the mating surfaces are near perfectly flat. While manufacturing techniques have certainly improved, it is not safe to assume the mating surfaces are flat enough, or that there are not too many microscopic pits and valleys. This is particularly true with aftermarket coolers - which Intel and AMD have no control over (and why using them violates the terms of their warranties).
• It assumes the user has properly mounted the HSF assembly, and it assumes the clamping mechanism is properly secured. While TIM does not positively ensure compensation for these mistakes every time, it does help ensure big gaps full of insulating air are not trapped between them.
• Temperature sensors are very cheap, low tech devices that can and do fail - and not all motherboards have them.
• PC Health settings in the BIOS Setup Menu can be disabled and should not be assumed to be working properly.
• While modern CPUs are supposed to shutdown for self-protection, it is not safe to assume they will.
• It is not correct at all to suggest you can run for an hour without TIM. There are any number of legitimate tasks that can instantly push a CPU to 100% utilization - especially if the OS sees the computer is "idle". And there are any number of illegitimate and/or malicious tasks that can push the CPU too. When a CPU is pushed to 100%, the temperature climbs rapidly, in seconds.
• The CPU fan can fail, or slow down significantly (keeping the fan sensor happy). This can especially be dangerous during testing when the motherboard is out of the case, or the case is open and there is no, or inadequate air flow from the case fans.

I do agree however, that using Arctic Silver for testing is expensive, and silly. So for testing in my shop, we use plain old, tried and true, white "goop" - silicone heat transfer compound, a staple of electronics repair facilities for decades. A standard tube of Arctic Silver 5 holds 3.5 grams of TIM and cost $13, and might supply 5 or 6 applications. A 2 ounce tub of silicone compound equals 56.7 grams and costs $3 less! It even comes in pint size tubs. Of course, plain old silicone compound is no where near as good as the modern TIMs, but it is much, much better than no TIM at all.

Running without TIM on your own computer is up to you and you do so at risk to your own wallet, though, IMO, that would be foolish. But when advising others, or when working on a client's machine, ALWAYS, as in EVERY SINGLE TIME, use a fresh layer of TIM. No exceptions.

A CPU should never be used without having Tim applied

That is absolutely correct and very sound advise.