GA-Z77X-UP5 TH » POST and reboot issues

[Haylee]

Hi,

I've been building a system with all-new components using the Z77X-UP5 TH in March-April of this year (system specs below). It is supposed to be a work machine with stability and reliability as the main focus, i.e. stock speeds and voltages if possible. It came with BIOS F11 and I am still using it as I was hoping for GB to post a more recent stable BIOS (which I now doubt they will ever do as the Z87 generation is now out). I've seen four issues with this build that have been causing me problems, which may or may not be related to one another:

- Debug LED D4 error code on cold boot, A0 on warm boot (PCI Resource allocation error, out of resources)
This has been noted here, here and interestingly also for UD3H and UD5H boards. Not sure if there are any negative consequences from this, as the system boots and appears to work normally either way. I have a GTX 670 in the first PCIe (x16) slot and a Soundblaster Z soundcard in one of the PCIe x1 slots, and they work fine regardless of what the LED code is. Not sure what to make of this other than I wish GB would offer a new stable BIOS that fixes it.

- Memtest errors when using memory sticks in XMP level 1 config
I noticed this right away as I was testing the mem sticks (2x8GB / 1600) with memtest86+ and IntelBurnTest during installation. At first I suspected that I simply had a bad pair of memory sticks, as just a single stick was failing. I requested an RMA, but the same thing happened with two more sets of memory sticks that Corsair sent me as replacements. There are always errors while in XMP mode, but less or no errors when at stock speeds (however some sticks would fail at stock speeds too!). I eventually received a pair from Corsair that would only fail when in XMP mode but would cause no errors after day-long runs of memtest86+ when at stock speeds. I was ok with that for the time being, pending a BIOS upgrade or switching to a different make of memory.. I haven't decided on what to do about that yet.

- Weird sporadic but very regular POST issue
This is being mentioned in one of the posts concerning the D4 error code as well: the system fails POST, only once each time, after about 16-20 boot cycles. I never got a precise number of cycles between each recurrence, but it seems to occur at very regular intervals. Obviously this came only to light after some time. As with the LED error code, it isn't exactly satisfying but it doesn't seem to cause any particular issues either: I am not losing work due to spontaneous reboots, just that single POST failure every 16-20 boot cycles or so. A weird BIOS bug, perhaps - and quite possibly related to the memory issue?

- A couple of months later: Spontaneous reboots and failure to POST
The most recent problem and the reason for my post now: suddenly the system began to reboot spontaneously, often some 30 minutes to an hour after boot-up, and not under load. Nothing to do with overheating or power loads, no BSODs or error messages, and I couldn't find anything in the windows event logs that would point at the cause. Driver issues seem unlikely to me since I hadn't updated any drivers recently (and disabled windows driver updates).

Also, POST failures (power cycle » POST failure » power cycle » eventually getting through to BIOS and OS after a couple of tries), pointing to hardware issues and making me suspect compatibility issues with the memory sticks again. Like all components in this build, the PSU (see below) is new and should be more than capable of handling this system.

However I just ran memtest86+ on these sticks (at stock speeds, not XMP) for a day without issues. Next, I am thinking of resetting CMOS and perhaps flashing the current beta BIOS F12j on it, as I see someone has had luck with this.

So now I wonder:
- Why do I see these problems now after the system seems to have been stable for several months?
- Can the BIOS get degraded/corrupted over time? (I didn't see any error messages to that effect though)
- What's the deal with the memory stick issues I am seeing? These are mem sticks from a leading manufacturer after all..
- Does the MB not like 2x8GB mem sticks? Should I try 4x4GB instead?

Thanks for your help,
Haylee


System Specs
MB           GA-Z77X-UP5 TH (rev 1.0)
BIOS        F11
Memory    Corsair Vengeance CMZ16GX3M2A1600C9 (2x8GB/1600 with XMP, 1333 stock speed)
CPU         i7-3770K
Cooler      Corsair Hydro H80i
GPU         Gigabyte NV670-OC2G (Nvidia GTX 670 2GB) in PCIe x16 slot 1
PSU         Corsair AX-760i
Sound      Soundblaster Z in PCIe x1 slot 1
Storage    OZC Vector 256G, WD Black, WD Red
Display     Dell U2713H
OS          Win 8 Pro 64bit

[ysw]

you may update to f12e to try or change other PSU to check.

[Haylee]

Thanks,
- are there any known compatibility issues with the Z77X-UP5 TH and Corsair AXi-type power supplies?

[ex58]

Thanks,
- are there any known compatibility issues with the Z77X-UP5 TH and Corsair AXi-type power supplies?

Yes,AXi series is not compatible with GB boards (noted on many Z87 series)........
Btw,
I suggest you to flash latest BIOS F12j on TT forum.....
http://forums.tweaktown.com/gigabyte/28441-gigabyte-latest-beta-bios.html

[Haylee]

After taking a closer look at the back side of the motherboard I discovered some sort of fluid seepage along the metal back plate that supports the heatsink between the CPU socket and the I/O ports. This doesn't look kosher to me.. you can see the fluid is both on the metal plate and also on the motherboard itself. It would seem this could easily cause an electrical short and make the PSU shut down the system.

Can somebody tell me if there's fluid in the heatpipe that connects the heatsinks on this board? I thought it was simply an air pipe. Perhaps one of the capacitors next to the ferrite chokes blew and is leaking fluid? I need to disassemble the build first before I can take a good look at the front, but most of that area is covered by the heatsink anyway.. In any case, there's some sort of fluid that would likely cause the MB to short out.

Fluid leaking along the metal back plate and seeping onto the motherboard as well:




Location of the metal back plate in reference to front side:

[Haylee]

Yes,AXi series is not compatible with GB boards (noted on many Z87 series)

Can you please expand on this? I am only aware of one specific issue
affecting the AX760i/AX860i and Gigabyte Z87X-OC and Z87X-OC Force mobos only ―
confirmed by Dinos (works for Gigabyte). It has to do with the way Ignition Mode
(a Gigabyte OC feature) is implemented on these boards.

But again, that is specific to those boards only. Are there similar issues with the AX760i/860i and Z77
series boards that you are aware of?

[dmdilks]

The leaking is more likely your problem. The board should still be under warranty. I myself would call or email Gigabyte.

Or you can get on line and setup a RMA with them. I just did it with a Ga-x79-UD3 board that had two bad memory slots.