Personal computing discussed

Moderators: mac_h8r1, Nemesis

 
DrDominodog51
Gerbil Elite
Topic Author
Posts: 639
Joined: Sat Jan 17, 2015 12:23 pm
Location: Silicon Valley

Infinitely failing to POST

Sat Mar 04, 2017 2:04 pm

Hey Gerbils,

The computer in my signature (A 4690k, Asrock z97 Pro4, 8x2 GB of DDR3, RM750, R9 Fury system) has started to consistently fail to POST and is power cycling itself infinitely to reattempt POSTing. The only way to stop the power cycle loop is to cut power to the machine. Given how many issues I've had with this motherboard in the past, I'm inclined to believe the motherboard has failed.

To get the computer to POST I've:

1) Cleared the CMOS

2) Removed the CMOS Battery after unplugging the machine.

3) Removed the GPU. This motherboard has had issues supplying power to the PCIe slots in the past and needed an auxiliary Molex to boot with the Fury in.

Are there an additional steps I'm missing to try fix the problem?

Thanks.
 
just brew it!
Gold subscriber
Administrator
Posts: 48540
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Infinitely failing to POST

Sat Mar 04, 2017 2:57 pm

Re-seat all RAM and power connectors? Try a spare PSU? Disconnect all storage devices (both internal and external) to make sure one of them isn't triggering a weird POST issue?
If the world isn't making sense to you, you're either drinking too much or not drinking enough.
 
Darthutos
Gerbil
Posts: 76
Joined: Sun Sep 07, 2014 7:39 am

Re: Infinitely failing to POST

Sat Mar 04, 2017 3:50 pm

Hi

You have said you removed the cmos battery. Do you mean you bought a new one and installed it or just removed it and then put the same battery later back later?

If the latter you need to buy a new cmos battery and try that.

Is there beeps or a debug display displaying weird numbers? like A2, or 01 or stuff on the mobo?
 
Kougar
Silver subscriber
Gerbil Jedi
Posts: 1854
Joined: Tue Dec 02, 2008 2:12 am
Location: Texas

Re: Infinitely failing to POST

Sat Mar 04, 2017 7:25 pm

A bad CMOS battery would cause a loss of settings, but not a boot loop.

Do everything JBI suggests. If that still doesn't work take the Fury out of the system, remove the molex plug from the motherboard then try booting to the IGP on your processor.
 
DrDominodog51
Gerbil Elite
Topic Author
Posts: 639
Joined: Sat Jan 17, 2015 12:23 pm
Location: Silicon Valley

Re: Infinitely failing to POST

Sat Mar 04, 2017 8:26 pm

Reseating the memory seems to have fixed the issue for now.

It still power cycles once upon boot due to the Fury, but it has always done that.

At this point, I'm going see if any replacement motherboards are still available at retail prices as this isn't my first issue with the motherboard, and I doubt it will be my last.

Thanks for the help everyone.

Edit: I found what was probably the issue. The default BIOS setting for memory training is set to skip it (and I didn't change this setting). Reseating the RAM probably causes the RAM to be trained.

I've changed the BIOS setting now, and I'll post back here if I have the same issue again.
 
just brew it!
Gold subscriber
Administrator
Posts: 48540
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Infinitely failing to POST

Sat Mar 04, 2017 9:13 pm

I can't imagine re-seating forcing a retrain unless one of the DIMMs had worked itself loose and wasn't being detected properly. The system doesn't have a way of knowing if you removed and reinstalled a DIMM while the power was off (unless you have different models of DIMM and moved them to different slots).
If the world isn't making sense to you, you're either drinking too much or not drinking enough.
 
DrDominodog51
Gerbil Elite
Topic Author
Posts: 639
Joined: Sat Jan 17, 2015 12:23 pm
Location: Silicon Valley

Re: Infinitely failing to POST

Sat Mar 04, 2017 10:16 pm

just brew it! wrote:
I can't imagine re-seating forcing a retrain unless one of the DIMMs had worked itself loose and wasn't being detected properly. The system doesn't have a way of knowing if you removed and reinstalled a DIMM while the power was off (unless you have different models of DIMM and moved them to different slots).

I have two different models of DIMMs with different timings and I believe I switched the channels they are on as well.

As an additional bonus, 102 MHz BCLK is stable now.
 
just brew it!
Gold subscriber
Administrator
Posts: 48540
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Infinitely failing to POST

Sat Mar 04, 2017 10:24 pm

DrDominodog51 wrote:
just brew it! wrote:
I can't imagine re-seating forcing a retrain unless one of the DIMMs had worked itself loose and wasn't being detected properly. The system doesn't have a way of knowing if you removed and reinstalled a DIMM while the power was off (unless you have different models of DIMM and moved them to different slots).

I have two different models of DIMMs with different timings and I believe I switched the channels they are on as well.

Ahh... yeah, that's an additional variable. That being the case, it is also possible that one of your memory channels is starting to go, and is just "happier" now with the other set of DIMMs on it.
If the world isn't making sense to you, you're either drinking too much or not drinking enough.
 
not@home
Gerbil
Posts: 24
Joined: Tue May 27, 2008 6:12 pm

Re: Infinitely failing to POST

Sun Mar 12, 2017 10:03 pm

I would check for any BIOS updates for the Mobo. Also I would test the RAM with Memtest x86+
 
I.S.T.
Gerbil XP
Posts: 486
Joined: Wed Sep 14, 2005 5:18 am

Re: Infinitely failing to POST

Sun Mar 12, 2017 10:30 pm

Memest ain't flawless, FWIW. Couple years back, I had some bad memory causing instability past about two to four days of uptime, and ran a full 72 hour Memtest. It didn't find ****. Eventually, I just tried replacing the memory, and what do you know but my system is rocksolid stable again outside of my own idiocy(Which crashes Ol' Faithful more than any other source).
 
just brew it!
Gold subscriber
Administrator
Posts: 48540
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Infinitely failing to POST

Mon Mar 13, 2017 4:45 am

I.S.T. wrote:
Memest ain't flawless, FWIW. Couple years back, I had some bad memory causing instability past about two to four days of uptime, and ran a full 72 hour Memtest. It didn't find ****. Eventually, I just tried replacing the memory, and what do you know but my system is rocksolid stable again outside of my own idiocy(Which crashes Ol' Faithful more than any other source).

Yeah, it doesn't find everything, but it does find most flaky DIMMs, IME. Back in the day, my SOP was to run Memtest86 overnight, followed by SiSoft Sandra's memory bandwidth test. Any flaky DIMMs that made it past Memtest86 would generally cause system freezes or BSODs under SiSoft Sandra.
If the world isn't making sense to you, you're either drinking too much or not drinking enough.
 
Kougar
Silver subscriber
Gerbil Jedi
Posts: 1854
Joined: Tue Dec 02, 2008 2:12 am
Location: Texas

Re: Infinitely failing to POST

Tue Mar 14, 2017 9:16 pm

just brew it! wrote:
DIMMs that made it past Memtest86 would generally cause system freezes or BSODs under SiSoft Sandra.


Quick question, would you consider SiSoft Sandra better than Prime95 blend runs for that purpose? I have the same SOP except with Prime95 Blend runs as those work faster than Memtest at finding RAM-related instability, but it's been a long time since I played with Sandra.
 
just brew it!
Gold subscriber
Administrator
Posts: 48540
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Infinitely failing to POST

Tue Mar 14, 2017 10:02 pm

Kougar wrote:
just brew it! wrote:
DIMMs that made it past Memtest86 would generally cause system freezes or BSODs under SiSoft Sandra.

Quick question, would you consider SiSoft Sandra better than Prime95 blend runs for that purpose? I have the same SOP except with Prime95 Blend runs as those work faster than Memtest at finding RAM-related instability, but it's been a long time since I played with Sandra.

TBH these days I really have no idea. Around the time DDR2 became a thing the incidence of flaky DIMMs seemed to go way down; the few bad DIMMs I've had since then have either been completely DOA, or Memtest86 has detected them. So I've basically stopped using anything beyond Memtest86 to test new RAM. I do still run several days of Folding@home as a burn-in test for new builds, but that's to test my CPU thermal solution (for both effectiveness and noise level under load); I don't think it particularly stresses the RAM.
If the world isn't making sense to you, you're either drinking too much or not drinking enough.
 
Kougar
Silver subscriber
Gerbil Jedi
Posts: 1854
Joined: Tue Dec 02, 2008 2:12 am
Location: Texas

Re: Infinitely failing to POST

Fri Mar 17, 2017 12:29 am

just brew it! wrote:
I do still run several days of Folding@home as a burn-in test for new builds, but that's to test my CPU thermal solution (for both effectiveness and noise level under load); I don't think it particularly stresses the RAM.


Prime95 is better for thermal testing, especially when it kicks on AVX runs. It'll alternate between AVX/non-AVX so it takes a few minutes before it will start running AVX loads. F@H is not a good checker for RAM, back when I was having all manner of RAM issues on two Haswell rigs it was always the F@H GPU client catching issues or the GPU driver itself, and ironically never the CPU client.

Who is online

Users browsing this forum: just brew it!, NTMBK and 9 guests