New parts, CPU LED red, PC won't post

MrAllfather

Honorable
Jan 11, 2013
23
0
10,510
Hey fellas, I am having some computer build issues, and you folks here at Tom's Hardware are always the most helpful and kind people when it comes to hardware issues. I have done heavy research, but have not read anything that has been able to fix my dilemma. People with similar problems rarely post a conclusion or solution after they figure it out. (Just a heads up, this post will be quite long, sorry!)

I will preface with current/new specs.
- Case: Rosewill Thor v2
- Mobo: m5a99fx pro 2.0
- CPU: AMD FX8350
- RAM: 2x8gb G.skill Ripjawx@2133
- GPU: AMD Radeon HD 7970
- PSU: EVGA G2 750watt

The situation began after I got back from a 2 week long vacation at the end of May. I booted my computer up (had been running fine for over a year) and after 15 minutes of internet browsing and downloading Witcher 3, it just shut off like someone had hit the PSU power. The case LEDs were blinking as if it were in sleep mode, but the power button did nothing. I unplugged the PSU and ran some diagnostics.

The power button did nothing even after reseating every component. I assumed it was either dead PSU or Mobo, so I bought the new M5A99FX Pro 2.0, the new Thor case, the EVGA PSU, and the new G.skill memory. Upon receiving these upgrades, I carefully put my new computer together (only old parts are cpu and gpu). Unfortunately, in my eagerness I skipped breadboarding.

The new issue was that the red cpu led on the board was always lit up, and the computer would not post. There is no signal detected by the monitor. However, all fans (cpu fan, case fans, gpu fans) and case leds were receiving power, and the green led on the mobo was lit - verifying a power flow. I thought the motherboard may have been DOA based on the merchant (newegg) feedback, and RMA'd the board for replacement.

I just got the new one but I have the same issue! Here are some things I have tried:
- PSU paperclip test (success)
- reseating everything and breadboarding
- cleaning every component thoroughly with soft paintbrush and compressed air
- different PCI and memory slot configs
- different monitor
- same monitor with different connector
- yes, CPU 8 pin was plugged in during all trials

Sadly I do not have any other parts to test with. I have not tried flashing BIOS because it was sent up-to-date and is compatible with the fx8350 already. I am just completely at a loss. In my experience, CPUs are very durable and usually the last component to have issues with if installed correctly and safely - yet I am beginning to think that the CPU may have been fried in my last build, as the old PSU was also dead (as per paperclip test) and the old Mobo did not work at all through breadboarding with the other new parts. Could this be the CPU?

Thanks so much for your time, and any input will be greatly appreciated!

EDIT: I forgot to mention that my reason for suspecting the CPU is that I get the same results powering on whether the EPU is plugged in or not. So it seems the CPU may be the source of the issue and is simply not being detected. There are some slightly bent pins, but I have combed the pins with a credit card just in case and everything seats properly and snug. I should also mention that this is a dedicated gaming pc that is used frequently for games that are graphically intense, and I have been using the 8350 and stock heatsink for about a year now... without reapplying any thermal paste (yikes, I know, never crossed my mind).
 
Solution
Chances of the CPU being the culprit are very, very low. I did repairs for a while and out of maybe 1000 CPUs that crossed my hands maybe 1 or 2 were bad, not counting bent pins of course, and those were normally bad because of extreme cases such as a bolt of lightning striking the home and melting everything. Hell one time a tech at a shop I was working at dropped a CPU and the small part of the PCB was chipped a bit (it was an Intel LGA chip so no pins to bend). I cleaned it off and tested it and it ran perfectly. Used it as a test chip for a few years.

The board needs a CPU and GPU to POST since it does not have an onboard GPU to run. The only thing that you have not tried that is a possible culprit is the GPU. You did say you don't...

napster100

Distinguished
Aug 13, 2013
401
0
18,860
Since you've changed the mobo, you'd have had to take the cooler off, since there was no new paste, that could have introduced air which of course is not good, but it does take a minute or 2 for the CPU to over heat so it should have POST successfully if that was the cause.

Have you tried a single stick of RAM and placed it in the slot furthest away from the CPU? If no POST try each slot. I had a problem the other day where my PC wouldn't POST because of RAM, it turns out the cooler was to tight, which was effecting the memory controller, you could try taking it off completely then placing back on with enough tightness to create contact, but with only a little bit of pressure.

Just to make sure, you did install the mobo standoffs on the case right? It happens aha.

It is possible the CPU is fried, since if the PSU caused a high voltage spike upon dying, it's possible. Bent pins on the CPU don't sound healthy at all, that itself could be the issue.

Do you have a speaker installed on the board and are you getting any beep codes?

Unfortunately trying onboard graphics isn't possible, because there is none... But I'd definitely recommend looking around the RAM and CPU area, the problem is likely there.
 
You don't need to reapply thermal paste often, normally once every couple of years and really depends on the thermal paste itself. The stock heatsink is fine if you are not overclocking.

A few things you need to do though:

- Check the standoffs you have, was your old motherboard the same size? Did it have the same standoff layout as the new one? If there is one in the wrong place touching the back of the board it could be shorting it out.

- Test without the GPU to see if it goes past the CPU LED

- Try with just a single RAM stick in the slot closest to the CPU

That's all I can think of. Might also want to try pulling parts out and putting them on a bench outside of the case that could cause issues.
 
It's pretty obvious something happened while you were on vacation. You said you rebooted upon returning. I'm guessing you powered off when leaving for vacation with the case power button and left everything connected and the PSU on. If the area had severe weather while you were gone it's possible that you had a lightning strike or several rapid power failures with quick restorations - say 3 to 5 in a couple of seconds. Whatever the case, it sounds like you had some kind of surge. If it was due to a nearby lightning strike the surge could have gotten into the machine via cables other than the power line.

You said you didn't breadboard the new build which kind of indicates you were in a hurry. Have you double checked that the CPU is installed in the socket correctly?

Bent pins can still be attached but have their internal connection broken due to being bent.

The only other thing I can think of is that someone else had access to the system and was using it while you were gone and they ran into a problem that you don't know about. The lack of thermal paste could play into this scenario.

With what you've posted, I'd say the CPU is toast. You might have a local shop check it for you if possible.
 

MrAllfather

Honorable
Jan 11, 2013
23
0
10,510
Hey guys, thanks for all the speedy and thorough responses. I am at work right now so I won't be able to test any suggestions until later, but I will adress some of the questions brought up.

The standoffs were one of the first things I checked. It is in ATX2 layout as per the motherboard manual, and nothing is too tight or too loose.

I have tried both memory sticks individually in every slot, as well as the old RAM.

While I was on vacation, I made sure to turn everything off and unplug the power strip.

I did breadboard after the mobo RMA, and the issue persists even without the GPU installed. Also, I have tried booting with the cpu fan loosely placed on the CPU in case it was too tight.

When I rebuilt the computer initially, I was not in a hurry, just wrongly confident that I had replaced the parts causing the original problem. I made sure to be well grounded and was careful handling the components.

I would take the computer into a local tech support shop if I had one. I live in the Rockies USA in a ski town, and the closest repair shop I know of via Google is a ~2 hour drive to Denver. There is probably a Best Buy or something similar closer, but I have had poor experiences with Geek Squad and the likes.

Thanks again for the responses and your time. I have no problem replacing the CPU, I just want to isolate the issue so that I don't have to deal with back and forth shipping or wasted cash.
 
Geek Squad is a bad place TBH. I have used their software and while you can use it properly and get good results, they don't know how to use it. That and they promote sales people to Geek Squad based on sales not knowledge.

The only thing I can suggest is try with just the PSU plugged into the mobo. Remove all other connections, even front panel connections and use a paperclip to jump the board for power. Don't even plug a USB mouse/keyboard in the back. Just the components.

I would say try without the GPU but the biggest problem is that that board does not support onboard graphics so no GPU would cause a no POST on it and if it is the GPU that is bad causing the no POST then without a proper test one we might be SoL.
 

MrAllfather

Honorable
Jan 11, 2013
23
0
10,510
@jimmysmitty

I will have to try that PSU test again when I get home. It turned on with the first paperclip test, but I can't recall what I had connected at the time.
Also, I have actually tried booting without the CPU or GPU connected, but I have the same problem. All fans run and power is on, but it just will not post. I figure it might be the CPU as it won't post without one and has the red cpu led whether it is installed or not. Almost as though it is not being detected.
Some additional info: I let the conputer run for 15 minutes and afterwards unplugged the power and felt the heatsink. It was slightly warm, but not enough to tell if the processor is even functioning.

Thanks.
 
Chances of the CPU being the culprit are very, very low. I did repairs for a while and out of maybe 1000 CPUs that crossed my hands maybe 1 or 2 were bad, not counting bent pins of course, and those were normally bad because of extreme cases such as a bolt of lightning striking the home and melting everything. Hell one time a tech at a shop I was working at dropped a CPU and the small part of the PCB was chipped a bit (it was an Intel LGA chip so no pins to bend). I cleaned it off and tested it and it ran perfectly. Used it as a test chip for a few years.

The board needs a CPU and GPU to POST since it does not have an onboard GPU to run. The only thing that you have not tried that is a possible culprit is the GPU. You did say you don't have one to test with but that could be the issue and you could be running around in circles. I would ask a friend if they have one they would be willing to let you use to just test with.
 
Solution

MrAllfather

Honorable
Jan 11, 2013
23
0
10,510
Thanks again, Jimmy.

Unfortunately the issue remains after redoing the PSU test as suggested. I know it won't post without CPU or GPU, but I also attempted booting without either of them to see if I got the same error led, and it yet persists.

EDIT: I suppose this issue just can't be solved by forum. I will just order a new CPU and GPU and pray it works. Though, if not, I can probably just return it for refund. Thanks for the input, everyone, I will post back here later when I have replaced the potential culprits so that anyone with a similar predicament can also reach a conclusion.
 

TRENDING THREADS