PC keeps rebooting after BIOS update

[stjepan.herceg]

This is just brutal... I had a PC with a Gigabyte Z87X-UD7-TH motherboard with a 4690k CPU and it just kept rebooting. I only ever saw the BIOS screen for a few seconds and then it restarted.

I opened the thread a week or so ago and we came to the conclusion that it must be a faulty motherboard. Now I've purchased a Gigabyte Z390 Designare with a 9900k CPU and Trident Z 3200Mhz 2x16 GB RAM and kept the GPU (780Ti), PSU(AX1200i) and SSDs.

After the hardware upgrade, I was a bit surprised how Windows 10 recognized the new hardware and began the installation of the new drivers automatically. Afterwards, I was able to use Windows 10 without any problems. First thing I did were the windows updates and the GPU updates as well as the installation of Gigabyte Apps from the CD. Afterwards, I tried to update the BIOS which was at version F3 when I started using it, and I did it with the @BIOS Gigabyte app. I downloaded The F6a firmware from their website and directed the @BIOS app, to use the downloaded and extracted BIOS file that was saved to my desktop. The process seemed unproblematic. It took about 2 minutes and afterwards, the app rebooted the PC. Before the reboot was complete, the PC shut down 2 times in a row before the bios screen was reached. On the next attempt, I saw the BIOS screen again, but it rebooted soon afterwards.

I then tried to clear the CMOS via battery removal (turned off PC, unplugged power cable, power button pressed for 60 seconds, battery removed, inserted back after 2 minutes, power on) and after the pc rebooted, the bios informed me, that the bios firmware was reset and I had to load optimized defaults or do it manually. I tried to do just that, but it rebooted again on its own.

The strange thing is that it still display F6a as the firmware version, even though it said, that the firmware has been updated.

Here are some picutures I took from the bios screen that I only saw for a few seconds:

Are some of these values off? Is vcore supposed to be around 1.2 and the speed at around 4.7 GHz in the BIOS? Is there anything else that I could look up in the few seconds I have left in the BIOS? What else can I try?

This has turned into a nightmare... unbelievable.. I hope the motherboard isn't bricked beyond repair already after just one day...

EDIT: I tried the CMOS reset via jumpers and after a few seconds I noticed a sort of lightning on the Mobo. Anyway, afterwards, the pc restarted but it was immediately cut off. This happened one more time again. And the third time, it rebooted as usual. I reached BIOS and I saw that F6a was still loaded... and it rebooted once again...

 

[SkyNetRising]

Try a different BIOS version - F5 for example.

BTW - how are you having cpu temperature at 19C? Is your room temperature like 10C?

 

[stjepan.herceg]

Try a different BIOS version - F5 for example.

BTW - how are you having cpu temperature at 19C? Is your room temperature like 10C?

Well I'd try a different BIOS version, but the system keeps rebooting. No matter what I do, after about 4-5 seconds, it reboots. I also tried to hold the power and reset button for 10 seconds and then I got a repair message on the gigabyte "ultra durable" loading screen, but it still shut down soon afterwards. Should I try it via Q-Flash / USB Stick?

The product page claims that the Z390 designare supports dual bios, but there's no real switch to choose between the first or second bios... Is there any way to use the second bios or is it a misleading info on gigabyte's page?

And yes, my room's pretty cold... I always keep my window wide open, and now it's winter...

 

[SkyNetRising]

Your board seems to have dual bios feature. Read manual, how to use it.
You should be able either to boot from secondary BIOS or
restore primary BIOS from secondary.

 

[SkyNetRising]

Hmm. Dual bios just allows to boot the system, if primary bios is corrupted.
You'll have to use q-flash to update bios.
But q-flash mode must be stable. If your pc is rebooting in q-flash mode, you won't be able to update bios successfully.

 

[stjepan.herceg]

Hmm. Dual bios just allows to boot the system, if primary bios is corrupted.
You'll have to use q-flash to update bios.
But q-flash mode must be stable. If your pc is rebooting in q-flash mode, you won't be able to update bios successfully.

The manual isn't exactly helpful in that regard. I can see that there is a B_BIOS and an M_BIOS, however, no explanations are given.

This is all I've found:

What is DualBIOS™?
Motherboards that support DualBIOS have two BIOS onboard, a main BIOS and a backup BIOS. Normally, the
system works on the main BIOS. However, if the main BIOS is corrupted or damaged, the backup BIOS will
take over on the next system boot to ensure normal system operation. For the sake of system safety, users
cannot update the backup BIOS manually.

The backup bios should be the F3 version, yet somehow I only get the F6a version even after CMOS resets, battery removals etc. Maybe F6a is not recognized as corrupted?

Unfortunately, there's no stable mode currently. Doesn't matter if I enter the system information menu, bios, proceed with windows boot up, q-flash menu etc. it always reboots after about 5-8 seconds.

 

[SkyNetRising]

Try with single ram module, no gpu, no drives and see, how that impacts stability.
May be try a different PSU, if available.

You can also try forced boot from secondary bios.
https://www.overclockers.com/forums/showthread.php/697533-GUIDE-Forcing-backup-BIOS-on-Gigabyte-motherboards

 

[stjepan.herceg]

Try with single ram module, no gpu, no drives and see, how that impacts stability.
May be try a different PSU, if available.

You can also try forced boot from secondary bios.
https://www.overclockers.com/forums/showthread.php/697533-GUIDE-Forcing-backup-BIOS-on-Gigabyte-motherboards

I already tried options 1 & 2 from the guide above and It didn't work. I did hear three quick clicks and it failed to boot two times in a row (no bios screen at all), but then the third time, it just booted into the usual bios.

If F6a is displayed in the bios menu, is it safe to say that the backup bios chip is not being used?

As for option 3: which bios chip do I need to use for this? B_BIOS or M_BIOS? He talks about red dots and pin 1 being placed just under the dot, but I only see a white dot on both my BIOS chips. Can I use thin insulated cable in order to bridge pin1 and pin6?

 

[SkyNetRising]

M_Bios - stands for main bios. B_Bios stands for backup. You'd have to short out contacts on main bios.
I must admit, I haven't personally done this procedure and have no idea, how safe it is. Damage may be possible, if done improperly. Make sure, you connect the right pins.

 

[stjepan.herceg]

M_Bios - stands for main bios. B_Bios stands for backup. You'd have to short out contacts on main bios.
I must admit, I haven't personally done this procedure and have no idea, how safe it is. Damage may be possible, if done improperly. Make sure, you connect the right pins.

Something very strange happened... It suddenly works... I turned it on and Windows started to load after a while. I then rebooted myself and loaded the optimized defaults, as I wasn't able to access internet sites, even though my wifi and ethernet connection should work. The firmware was kept at F6a, the version which I used in @BIOS tool, so the backup bios did not "rescue" the main bios. The only problem that I have now is that the internet is still not working (firefox and other browsers are prohibiting sites due to certificates), even with optimized defaults. It did work before, as I was able to install all the windows, kaspersky updates etc.

Is this due to the omittance of a clean install of windows 10? Or is the BIOS itself still corrupted?

***EDIT***: I'm back at square one. Once I enabled XMP Profile 1 and adjusted a couple Thunderbolt related bios settings, the pc is once again stuck in reboot loop. The successful attempt I had earlier on was probably due to leaving the psu off with the power cable connected for 12h+. I'll wait once again and see what happens. The internet was blocked by my anti-virus, as the system time was displayed to be the year 2098. Once I fixed the date, the licenses started working again and I was able to make the internet work. If only I had flashed the system once again when I had the chance...

 

[SkyNetRising]

XMP might be the culprit there.

 

[stjepan.herceg]

XMP might be the culprit there.

I did the clear CMOS thing afterwards, and this hasn't changed anything at all, so I guess it's mainly about not being able to change anything in the BIOS without it leading to incessant reboots. Maybe this MB is not compatible with the F6a version..

The trick from earlier on worked once again: I left the pc alone for the last 12 hours or so and powered it up again. At first, it rebooted immediately after just 1 second or so. Then Windows was loaded properly and I could use all the functions once again without any reboots.

I then formatted a USB 3 32GB stick to FAT32 (16KB sizeblocks), loaded the F6a firmware once again from the Gigabyte website, went into Q-Flash menu, let it update the firmware and the the reboot timer went on. Afterwards, the reboot did happen, but only for a few seconds. The screen showed nothing and I could not reach the BIOS.

After a second reboot, I went into the BIOS, but then the reboot cycle once again started to occur...

Here's what I don't get:

What does leaving the PSU turned off (with the power cord still plugged in) for about 12 hours (I tried it after just 30 minutes and it didn't work, so it has to be a longer time period) do to the motherboard that let's it start properly?

This is soo strange...Can I emulate this with a longer "clear CMOS via jumper" method? Or can I remove the battery for a longer time to get the same effect?

If I once again have to wait half a day, I'll make sure to use a different bios version with q-flash (F5 probably) the next time I get the chance.

 

[SkyNetRising]

12h period is most likely for capacitors to fully discharge.
You can get the same effect, by unplugging the pc from power and holding down power on button for several seconds.

 

[stjepan.herceg]

12h period is most likely for capacitors to fully discharge.
You can get the same effect, by unplugging the pc from power and holding down power on button for several seconds.

I turned off the PSU, I then unplugged the power cable and I pressed the power button for 60 seconds, then I pressed it again a couple of times. Plugged the power cable in, turned on the PSU and started the system. It did not yield the same results as the 12h rest. Instead I got the reboot loop.

Should I try to press the power button after just switching off the PSU but keeping the power cable connected to the PSU? Would this damage the PSU or the motherboard? After all, I left the PC for 12h with PSU switch turned to off and the power cable plugged in. Maybe I wouldn't have gotten the same results if I would have removed the power cable during this process.

 

[karenjoly]

The things I notice are the several unintentional oc's going on here. What do you think ? Should the dram freq be 2133.33 mhz ? That's 4266 DDR. No memory works that high, maybe no system boots that module as well.

And then there is the processor speed-4700.97 MHz. Maximum single core turbo boost is 5 GHz and given the 8 cores , 4.7 GHz suggests the cpu is on full boost on all cores from the get go even though no stress requires turbo boost at this time. Then I need not mention that the bios ver is misreported.

What are the chances that two independently manufactured items, the cpu and the memory module, will fire up out of spec on their own ? Zilch I think.

So, .......you have my deepest sympathy.

 

[stjepan.herceg]

The things I notice are the several unintentional oc's going on here. What do you think ? Should the dram freq be 2133.33 mhz ? That's 4266 DDR. No memory works that high, maybe no system boots that module as well.

And then there is the processor speed-4700.97 MHz. Maximum single core turbo boost is 5 GHz and given the 8 cores , 4.7 GHz suggests the cpu is on full boost on all cores from the get go even though no stress requires turbo boost at this time. Then I need not mention that the bios ver is misreported.

What are the chances that two independently manufactured items, the cpu and the memory module, will fire up out of spec on their own ? Zilch I think.

So, .......you have my deepest sympathy.

No, Karen, I think the DRAM frequency should not exceed the factory recommended 3200MHz (1600MHz). The spec sheet shows that the Z390 Designare is able to deal with 4266MHz and beyond, but somehow I'm a bit skeptical...

I also believe that the CPU shouldn't fire out of all cyllinders during menial tasks, yet somehow it does.

How is the BIOS version misreported? What version should be shown instead? I had pictures taken with the backup bios active (F3) and the main bios (F6a).

In conclusion, is it safe to say that the BIOS is corrupted or is the motherboard already damaged beyond repair? It's hard to interpret feelings such as sympathy over the internet...

 

[karenjoly]

Ok about the bios versions, just so you noticed the change.
There is no reason for the memory to post that high, way higher than the module rating and with XMP disabled or for the cpu performance. Together these are the reasons the board is in trouble but whether a new bios can fix it, I do not know.You may be able to RMA the board. If you need another board, try ASUS. I have had no problems with that brand.

 

[stjepan.herceg]

Ok about the bios versions, just so you noticed the change.
There is no reason for the memory to post that high, way higher than the module rating and with XMP disabled or for the cpu performance. Together these are the reasons the board is in trouble but whether a new bios can fix it, I do not know.You may be able to RMA the board. If you need another board, try ASUS. I have had no problems with that brand.

After the 12h wait, I was able to change some settings, and when I chose the first XMP profile, it showed 3200MHz for DRAM frequency. If 2133 MHz corresponds to 4266 MHz, wouldn't this imply that the XMP profile is also corrupted?

What I don't get is why this board is putting out these numbers with the backup bios version. Or maybe I'm still not reaching the bakup bios, but a temporary backup bios flashed main bios? I don't know... it's very weird...

I would have chosen Asus from the get-go, but unfortunately, I'm doing a hackintosh build and Gigagbyte boards are strongly advised in the hackintosh communities.

 

[SkyNetRising]

Ram frequency shown in BIOS is effective DDR frequency (2x).
So if it shows 2133mhz, then real frequency is (2 times less) 1066mhz.

And no - your 3200mhz memory will not operate at 4266mhz.

 

[karenjoly]

Re the DRAM frequency . As reported @ 2133 Mhz, it is more likely to be the actual DDR rate operating at that rate by default. When you set the XMP for 3200 MHz , the rate for the memory you bought, that should be reported as 3200 MHz. Did that rate show up and you disabled XMP at some point ? If the shown rate is just the default memory speed then the default is not a problem.

 

[stjepan.herceg]

Re the DRAM frequency . As reported @ 2133 Mhz, it is more likely to be the actual DDR rate operating at that rate by default. When you set the XMP for 3200 MHz , the rate for the memory you bought, that should be reported as 3200 MHz. Did that rate show up and you disabled XMP at some point ? If the shown rate is just the default memory speed then the default is not a problem.

After a 12h wait, I can get the PC to boot up in a stable way. The BIOS is of course not modified. I enter to BIOS to make these changes, e.g. set XMP to profile 1 and make a couple of other hackintosh related changes, and once I save & exit, the pc reboots and then gets stuck in the reboot loop once again. If I leave the BIOS options as they are after the 12h, I can restart and shut down windows 10 as many times as I want. However, if I make some changes to the BIOS, it's back to square one...

And yes, when I choose XMP profile 1, I get the automatic 3200MHz value in the BIOS, so I guess it's just the actual RAM frequency value, instead of the doubled value.

 

[stjepan.herceg]

I waited another 12h hours just to try a different bios firmware version (F5):

As you can see, the godforsaken q-flash program had some visual tricks up its sleeve... luckily for me, the update process reached 100% and the reboot was initiated. Afterwards, the PC tried to start (no bios screen came up), and it was rebooted once again. Then I went into the BIOS, applied optimized settings, saved & exited only to be stuck in the reboot loop once again...

And now another 12h wait until I can do anything...