Intermittent issues, motherboard? Looking for 2nd opinion.

M

manns41078

Contributor
Joined
Oct 10, 2013
Posts
16
Hi everyone.
I installed a new SSD 4 weeks ago.
2 days after installing it I installed Windows 10. Everything was awesome for a couple more days. Then when I rebooted the BIOS slowed to a crawl. The machine would still eventually boot, but POST and subsequent BIOS tests were taking around 3 minutes. Even the BIOS menus were slow to respond (also there is an animated bar at the top of the bios menu screens that was moving super slow, so it doesn't appear to be an input device issue). I finally unplugged everything and started plugging things back in. Sometimes it would be fine, and sometimes I would add a component and it would not be fine on the first boot after adding the component, but was OK on subsequent boots. Finally I reconnected everything and it has been fine for a couple of weeks. Note that I have tried removing the new SSD since it was the most likely culprit, but the BIOS is still exhibiting this behavior without the new SSD.
I should also mention that I checked the CMOS battery and its voltage was a bit low so I replaced it.

So everything was fine for a couple weeks, but then yesterday the PC failed to hibernate and I had to power down with the power button. Note that it had rebooted, soft and hard, multiple times over the past few weeks. Well this time The BIOS said disk read error. Then after another reboot it was going to a grub recovery prompt and I don't believe I have ever installed Linux on this system. The BIOS has a little OS in it so I assume that the grub recovery prompt was for that and not my boot drive. I was able to get to a boot device selection menu just fine and boot from CD or USB. However the BIOS was still horribly slow to respond to input.

So I started running tests. Drives all passed no problems. Memtest found a single bad sector in RAM so I resolved to test more. I reboot to try my normal boot drive and... No problems... Windows 10 boots right up, everything is awesome. Later on I reboot again after getting a newer version of memtest to try out. BIOS stops dead. It gets to the post beep, but the screen freezes at a logo before any detection tests begin. Crap. I power down and turn off the PSU For a few minutes. Boot back up and everything is alright but the super slow BIOS remains.

So I went back in to boot via USB to a win pce environment and ran more tests. Hard drives still ok, malwarebytes found 3 pups so I removed them. For the heck of it I said, "maybe the BIOS was infected somehow..." So I got a fresh copy of the ROM and flashed it. Well... No real change.

I can boot successfully most of the time. But the BIOS remains slow to respond.
BIOS reports all voltages OK.
I removed all but one RAM chip to test individually. First one was fine (BIOS still slow). Second one is in solo and BIOS is still slow.

So at this point I'm fairly convinced there's something wrong with the motherboard. I've always been a bit suspicious of it since it takes 10 seconds to do the POST beep. Always has over 2 different PSUs.

The MB is asus m4a88td-v Evo/ usb 3
RAM is 4x4GB of ddr3 at 1333mhz by Corsair.
New SSD is a Sandisk 980GB.
PSU is an Ultra X4 at 750w.
2 x Ati Radeon 5850 in Crossfire (yes I've tried removing them and using the onboard video)

So, what do you guys think?
Thanks!
 
(also there is an animated bar at the top of the bios menu screens that was moving super slow, so it doesn't appear to be an input device issue).
Click to expand...
Not an "input" device but failing to move through the BIOS during boot does suggest hardware.

Even 1 error with MemTest is bad. If you haven't already, you need to re-test with only 1 or 2 sticks, then swap in the others until, through a process of elimination, find the bad stick. Note that no software based RAM tester is conclusive. They can and often do report no errors but the RAM may still be faulty. HOWEVER if they report errors, this is a problem. RAM can test fine when tested or run alone but fails when paired with other sticks.

I've always been a bit suspicious of it since it takes 10 seconds to do the POST beep.
Click to expand...
??? 10 seconds is not slow. It is not quick, but not slow. What are you expecting?

If you disconnect your drives, how long does it take until you get a disk not found error? Since seeking a boot disk is at the very end of POST, this should give you an idea if this is really a BIOS/Motherboard issue or a drive/driver issue. I would try this with both your graphics cards, 1 card, and onboard.

You might also disconnect your network too. I have seen something as simple as a bad Ethernet cable (or router port) cause similar boot delay problems. Lost several chunks of hair over that! ;)
 
RE: the power button to POST beep time - my older computer was under 5 seconds. I guess I was always disappointed with the slow POST on this board. If you say 10 sec is alright then I'm not too worried about it. It has just always made me a little suspicious of the board's quality.

Well, I started again from scratch removing and re-connecting various components. Seems I'm battling multiple issues:
  1. My powered USB hub (or a device connected to it) seems to cause the issue with the slow BIOS boot and slow-to-respond BIOS menus. With it connected boot takes 60 seconds from power to no disk found - BIOS menus are slow to respond. Disconnected it takes 30 seconds - BIOS menus respond normally.
  2. One of the SATA ports on the motherboard appears to be faulty. Any drive connected to it and the BIOS fails to find any bootable drive on first boot after disconnect. Subsequent boots result in a boot loop and I can't even get into the BIOS to see if I can reconfigure anything. I've tested with multiple SATA cables and multiple drives. It seems this port is no good and probably the culprit of my previous boot issue and maybe even the boot into grub recovery.
  3. So far 2 RAM chips check out OK when tested solo. It takes around 2 hours to test a single chip, I have 4. When I tested all 4 simultaneously the test took 14 hours. Hopefully the bad chip will show itself when tested individually, otherwise I could spend the rest of the year trying to determine the bad chip.

So... I'm getting somewhere. Thanks for your reply and making me rethink about going through and testing peripherals yet again. However, if I'm getting intermittent issues with a USB peripheral AND an SATA port do we think the motherboard is breaking down? The board is 5 years old now and has been in Florida for 3 years (though the house has central air and is pretty dry). And I work from home so the PC is running around 16 hours/day 7 days/week, and I usually only set it to hibernate. No visible corrosion and is kept pretty clean. So while I've found some issues I'm still a bit suspicious of the motherboard.

Anyway, thanks, I'd love to hear any further thoughts.
I'm off to continue memtesting... yay!
 
Ok. So thankfully, after disconnecting from the bad SATA port and the USB hub, memtest takes about 20minutes per chip. The third chip I tested had 4 errors. It is a Corsair with lifetime warranty so it's easy to replace.
I may spend $100 and just get a new mb for peace of mind, but at least now I think
I'm back to a stable system.
Running another memtest now with the remaining 3 ram chips. Hopefully they're all ok together.
 
Hi manns41078. :smile9:
Curiosity: what memtest are you using? (HCI design/memtest 4.1, memtest.com/PassMark memtest86 6.2.0, memtest.org/memtest86+ 5.0.1)
 
Passmark, but with my motherboard it would only run v4.

The remaining 3 chips passed combined testing for 3 passes. I'm feeling pretty safe with them now. Hopefully no more issues until I can pony up for a newer MB.
Chip #4 is being shipped back to Corsair on Monday. It had a good run!
 

Has Sysnative Forums helped you? Please consider donating to help us support the site!

Back
Top