The dreaded CPU LED light (Asus Sabertooth 990FX R2.0)

[mcpullm]

Last week, my computer shut down while in "sleep" mode, and when I tried to turn it back on, I got the red CPU LED light and no POST. Specs were as follows:

Motherboard: Asus Sabertooth 990FX R2.0
Processor: AMD FX-8350 w/stock Wraith cooler
RAM: Team Elite 4GB DDR3-1333 CL9 x4
GPU: XFX Radeon RX 580 8 GB GTR XXX
SSD: Samsung 860 Evo 1 TB 2.5" Solid State Drive
PSU: Cooler Master Elite V2 - 550W

Attempting to narrow down the problem, I have replaced the motherboard (same model), processor (same model), and PSU (EVGA BQ 600 W 80+ Bronze Certified Semi-modular ATX). I also removed all the RAM and tried to boot with just one stick, swapping each one out. The machine still will not POST, and the red CPU LED continues to mock me.

Other info that might be relevant:
I tried replacing the CMOS battery on the old motherboard; no change. I have not tried it on the new board.
I tried clearing CMOS on the old motherboard; no change. I have not tried it on the new board.
All of my case fans and the Wraith cooler have continued to run when I try to boot.
The GPU fans do not run when I try to boot, but the VGA light does not come on. There was no indication of problems with the card prior to the crash.

Any help would be appreciated.

 

[Deleted member 24505]

Was the new board replaced with a verified working one?

 

[P4-630]

Any overclocks with the new hardware?
If so, try everything at stock speeds first.

 

[eidairaman1]

Sounds like you had a cpu die.

That led is a good indication of that.

 

[mcpullm]

Answering like this because I don't understand how the multiquote works:

Yes, the board was verified working.

I've never overclocked anything, and wouldn't know how to start trying, especially with a system that won't boot.

It's entirely possible the old CPU died, but it's not working with the new one, either.

 

[MachineLearning]

I know you've tried different sticks of the same kit, but have you tried entirely new RAM? I've had the CPU light go off, when in reality the memory was at fault.

 

[mcpullm]

I know you've tried different sticks of the same kit, but have you tried entirely new RAM? I've had the CPU light go off, when in reality the memory was at fault.

That's one of my next possible steps. I'm posting here to see if anyone has any better ideas than I can come up with.

 

[Bones]

Last week, my computer shut down while in "sleep" mode, and when I tried to turn it back on, I got the red CPU LED light and no POST. Specs were as follows:

Motherboard: Asus Sabertooth 990FX R2.0
Processor: AMD FX-8350 w/stock Wraith cooler
RAM: Team Elite 4GB DDR3-1333 CL9 x4
GPU: XFX Radeon RX 580 8 GB GTR XXX
SSD: Samsung 860 Evo 1 TB 2.5" Solid State Drive
PSU: Cooler Master Elite V2 - 550W

Attempting to narrow down the problem, I have replaced the motherboard (same model), processor (same model), and PSU (EVGA BQ 600 W 80+ Bronze Certified Semi-modular ATX). I also removed all the RAM and tried to boot with just one stick, swapping each one out. The machine still will not POST, and the red CPU LED continues to mock me.

Other info that might be relevant:
I tried replacing the CMOS battery on the old motherboard; no change. I have not tried it on the new board.
I tried clearing CMOS on the old motherboard; no change. I have not tried it on the new board.
All of my case fans and the Wraith cooler have continued to run when I try to boot.
The GPU fans do not run when I try to boot, but the VGA light does not come on. There was no indication of problems with the card prior to the crash.

Any help would be appreciated.

Lets look at what's common and what's not between the attempts you made to get it going again, specifically in terms of hardware.

Same: RAM plus the GPU used.
Different: Board and PSU in use along with the CPU.

Symptoms:
Red CPU LED coming on and staying on.

So, it seems the CPU itself isn't starting up inspite of a different PSU, board and CPU's in use.
That's why I was sorting what was common and not between your attempts to fix it. It's not unusual for the GPU fan to stay off during POST, most if not all these days are supposed to do that anyway, only starting up when they reach a given temp. It's something I personally don't like it but that's how they are so, unless you've seen it act differently while it was working just know that tends to be normal.

The red LED can be caused by defective RAM or RAM settings not right.
Be sure to check your CMOS battery voltage with a voltmeter, if it's below 2.9v's replace it, if it's above 2.9v's it's probrably OK but if right at the limit (2.91V's) replace it anyway because you will before long.
Be sure to attempt to clear the BIOS, by disconnecting/killing power from the wall and then remove the battery, then press the power button a few times.
Leave it like that for about 10 minutes, then pop the battery back in first and then restore power to try it again.

If there is an issue with your RAM settings you can force a default boot by pressing and holding the power button until it cuts off on it's own. You can do this three or four times in a row, back to back and then see if it will boot at all. You will get the error at POST prompting you to go in to the BIOS (Press F1) but that's normal to see whenever you do this and it defaults.

I'd also suggest not connecting anything not required to make it boot, including all drives because a bad drive can cause it to hang while booting. All you need is your keyboard to be hooked up where it should be, the mouse itself isn't required to go for a basic "Will it work or not" kind of boot.

From what I"m seeing by what's described it could be a bad GPU, a bad stick of RAM.
I will mention just because the GPU showed no signs of trouble before doesn't mean it's still good NOW, everything when it fails has that exact point in time failure occurs and it goes from there.

I'd try it with a basic test setup as described and try different things like a different GPU, I'm thinking that may be your problem but then again, it could be anything at this point.... And don't be afraid to test the unconventional as well.

 

[eidairaman1]

Answering like this because I don't understand how the multiquote works:

Yes, the board was verified working.

I've never overclocked anything, and wouldn't know how to start trying, especially with a system that won't boot.

It's entirely possible the old CPU died, but it's not working with the new one, either.

Well that indicates cpu dead then

Lets look at what's common and what's not between the attempts you made to get it going again, specifically in terms of hardware.

Same: RAM plus the GPU used.
Different: Board and PSU in use along with the CPU.

Symptoms:
Red CPU LED coming on and staying on.

So, it seems the CPU itself isn't starting up inspite of a different PSU, board and CPU's in use.
That's why I was sorting what was common and not between your attempts to fix it. It's not unusual for the GPU fan to stay off during POST, most if not all these days are supposed to do that anyway, only starting up when they reach a given temp. It's something I personally don't like it but that's how they are so, unless you've seen it act differently while it was working just know that tends to be normal.

The red LED can be caused by defective RAM or RAM settings not right.
Be sure to check your CMOS battery voltage with a voltmeter, if it's below 2.9v's replace it, if it's above 2.9v's it's probrably OK but if right at the limit (2.91V's) replace it anyway because you will before long.
Be sure to attempt to clear the BIOS, by disconnecting/killing power from the wall and then remove the battery, then press the power button a few times.
Leave it like that for about 10 minutes, then pop the battery back in first and then restore power to try it again.

If there is an issue with your RAM settings you can force a default boot by pressing and holding the power button until it cuts off on it's own. You can do this three or four times in a row, back to back and then see if it will boot at all. You will get the error at POST prompting you to go in to the BIOS (Press F1) but that's normal to see whenever you do this and it defaults.

I'd also suggest not connecting anything not required to make it boot, including all drives because a bad drive can cause it to hang while booting. All you need is your keyboard to be hooked up where it should be, the mouse itself isn't required to go for a basic "Will it work or not" kind of boot.

From what I"m seeing by what's described it could be a bad GPU, a bad stick of RAM.
I will mention just because the GPU showed no signs of trouble before doesn't mean it's still good NOW, everything when it fails has that exact point in time failure occurs and it goes from there.

I'd try it with a basic test setup as described and try different things like a different GPU, I'm thinking that may be your problem but then again, it could be anything at this point.... And don't be afraid to test the unconventional as well.

Wouldnt the ram led come on then?

 

[dont whant to set it"']

Wouldnt the ram led come on then?

When hardware goes into such limbo it could be anything.
To name a few: improperly seated cpu, mb only posts when the cpu cooler retaining mechanism is tightend "engough" .

 

[Bones]

Wouldnt the ram led come on then?

Not always, the CPU LED has to come on and go out first, then the RAM LED will light and do the same as part of the booting process as the machine goes down it's checklist of things to do and look for before POST.
If the CPU LED never goes out that means something is stopping it from proceeding further into the boot sequence - Most likely a dead CPU but in this case I'm not so sure since the OP did state they tried two different CPUs in different boards with an identical result.

I simply took what was common from all the testing done and made mention of it.
ATM it looks like the same RAM, GPU and drives were used with each attempt - The board CPU, and PSU being different as testing progressed with the same result from it each time.

 

[mcpullm]

Well, I don't think it's the RAM. Tried four all-new sticks, alone and together. Next try: The GPU.

 

[Mr.Scott]

It's the drive.

 

[MachineLearning]

It's the drive.

My American Megatrends ASUS boards hate new SSDs when operating in IDE SATA mode (SK Hynix Gold S31, not MX500 for example). It completely freezes up on boot. If the 860 Evo is one of these drives, and BIOS was somehow reset - yep, checks out

 

[eidairaman1]

My American Megatrends ASUS boards hate new SSDs when operating in IDE SATA mode (SK Hynix Gold S31, not MX500 for example). It completely freezes up on boot. If the 860 Evo is one of these drives, and BIOS was somehow reset - yep, checks out

Im using an 840 Pro

 

[mcpullm]

My American Megatrends ASUS boards hate new SSDs when operating in IDE SATA mode (SK Hynix Gold S31, not MX500 for example). It completely freezes up on boot. If the 860 Evo is one of these drives, and BIOS was somehow reset - yep, checks out

If this is the case, what's the solution?

 

[AlwaysHope]

Got an 850 Pro working well with my Sabbertooth 990FX R2.0 board.

 

[Bones]

If this is the case, what's the solution?

First:
See if the system will at least POST with all drives disconnected. Do use the same RAM and GPU when doing this so you can see if they are still in working order - If it fails to POST then you can rule out the drive(s) for now at least.

If it does POST that zeros in on the drive(s) and from that point it's just a matter of figuring out which drive is causing the problem.

A corrupted OS CAN cause things to not work so if you don't need anything from it, just erase/zero it out and see if the machine will POST with it in place and connected after the wipe.
If so, that means a corrupted OS was causing it but if not, that means the drive itself is faulty and needs replacment.

However, be aware of this too:
Beware of an optical drive if you have any in use causing it too.
If the machine will POST with the primary HDD/SDD you can start by hooking up other drives, one drive at a time until it fails to POST.
I'd also beware of a bad IDE/SATA cable too, that can cause problems as well.

If it's drive related, once you zero in on which one it is try a different IDE/SATA cable before ruling that particular drive as "Bad".

 

[MachineLearning]

If this is the case, what's the solution?

I'd follow Bones's procedure first above to diagnose if it is indeed a problem.

If we assume it is the issue, then what I do to get it to simply boot: try disconnecting every SATA device. If it boots, go into OS and set onboard SATA to AHCI mode.
JMicron ports (white & orange IIRC?) will do fine with newer drives as an alternative but I don't recommend that.

Im using an 840 Pro

It's not an issue for every new drive, nor every older ASUS board, and only in IDE mode (default setting upon CMOS clr).

 

[eidairaman1]

I'd follow Bones's procedure first above to diagnose if it is indeed a problem.

If we assume it is the issue, then what I do to get it to simply boot: try disconnecting every SATA device. If it boots, go into OS and set onboard SATA to AHCI mode.
JMicron ports (white & orange IIRC?) will do fine with newer drives as an alternative but I don't recommend that.


It's not an issue for every new drive, nor every older ASUS board, and only in IDE mode (default setting upon CMOS clr).

Yeah I just made sure my stuff is in acpi/apic from the days of the Athlon XP.

 

[mcpullm]

First:
See if the system will at least POST with all drives disconnected. Do use the same RAM and GPU when doing this so you can see if they are still in working order - If it fails to POST then you can rule out the drive(s) for now at least.

If it does POST that zeros in on the drive(s) and from that point it's just a matter of figuring out which drive is causing the problem.

Just tried this with both my SDD and my optical drive disconnected (unplugged SATA cables from the motherboard). Still fails to POST.

 

[xtreemchaos]

just thought id throw this in mounting pressure some not all FX CPUs can suffer from this and chuck out errors i came across this with a fx8350 and a 970 gaming msi board last year when chucking together a movie feeder for my son. dont know how you could adjust the pressure with your cooler tho.

 

[Bones]

Just tried this with both my SDD and my optical drive disconnected (unplugged SATA cables from the motherboard). Still fails to POST.

Then the next step is to try a different GPU here.
Doesn't have to be a higher/newer model, just one that works with the board for use of confirming it working or not.
As to this:

just thought id throw this in mounting pressure some not all FX CPUs can suffer from this and chuck out errors i came across this with a fx8350 and a 970 gaming msi board last year when chucking together a movie feeder for my son. dont know how you could adjust the pressure with your cooler tho.

This is a possibility in itself, if the cooler you are using is somehow messing with how the CPU is interacting/interfacing with the socket you'll have to solve that problem.
Uneven pressure on the socket can cause headaches like this. Be sure the cooler isn't trying to sideload it's mounting pressure to one side due to a bad/defective mounting system or for some other reason, including (I must say this) "User Error" and don't think I haven't been guilty of that before because I have.

 

[xtreemchaos]

i put it down to some AM3/+ sockets being a pain in the ass .

 

[natr0n]

Monitor maybe ?