Question Almost New Build Working Perfectly But Not Now

[Mantrid-Drone]

I have been putting the finishing touches to a long gestated Win7 build this week:-

Win7 Pro SP1 64bit
i5-3550
16GB (Corsair XMS3 ddr3 1600 4x4) matched
GA-Z77-DS3H MB
Crucial 120GB SSD
Seagate ITB HDD
WD 160 GB additional HDD
MSI HD 7850 1GB GPU

All hardware working fine so I installed the OS from flash drive, no problems. Driver install + updates no problems so I went ahead and started installing the programs I wanted, security software and personilized some setings and that too went perfectly. PC booting without issue and everything working as you'd hope.

So last thing to do: Windows updates. I'd slipstreamed the Convenience Update Rollup into the installer and that and the required earlier one was listed in the updates. All as expected. Checked for new updates: 20 'important' and 18 optional. I just went for all the 'important' ones and started the downloading. Just under 500MB in total and then the installing went ahead. All was fine until 14 out 20 when the Monthly Malicious Software Removal tool August (2019) took a hell of long time to install.

I left it for an hour - still no progress so I stopped the install using the button provided. But still the Windows Update process continued..........I left it for 5 minutes before shutting it down with Task Manager.

The PC went back to desktop; I did a few other things like look up what to do if Windows Updates fail, decided I'd try updating Windows again tomorrow and then closed down the PC in the normal way. Only thing out of the ordinary is that I noticed there were some Windows updates queued for install on shutdown. So as it shut down I had the usual Windows is configuring message, normal in such circumstances but otherwise it was a unremarkable shutdown.

The problem is that when I went to restart the PC about 30 minutes later it did not give the usual POST bleep, the MB logo didn't appear and the screen was blank. I then noticed the internet connection was not showing on the router and to cut a long story short: the fans are running fine, the HDDs are clearly spinning up but there is no output at all. Keyboard and mouse have no power, screen is recceiving no output from the GPU and the two flash drives I have attached - no power.

You can't launch the MB BIOS/UEFI and obviously there's no way to boot in Safe Mode. I even tried one of the Windows install discs I have - nothing.

So what the hell has gone wrong? Surely It can't be just a coincidence a previously 100% working new PC shuts down after a failed Windows update and then this happens.

But what could it be? I'd appreciate some help with this because I have not the faintest idea even where to start to try and sort this out.

 

[Flayed]

Maybe it's Microsoft's new tactic to increase Windows 10 adoption rates?

 

[VirtualLarry]

I heard rumors once, about a "Windows Update" that flashed the BIOS, supposedly to help patch Meltdown / Spectre. (Chances are that I heard wrong, and it was really just a microcode update patch in Windows. But you never know. Sometimes OEMs push things down on branded boxes, that are OEM-specific, and could in fact include BIOS flashers / updaters. Some, like HP, do this using their own Update tool.) (Edit: Possible, but that wouldn't prevent you from using a hotkey to access BIOS. Unless, your BIOS was configured for Fast Boot before you installed Windows. Sometimes, that's the BIOS default. Once you install Windows, you can't use a hotkey to access BIOS, when Fast Boot is Enabled, and/or Secure Boot, unless you disconnect the OS drive physically, and try to boot without it, I've found.)

There was also another one, that leads to a "Red Screen" upon boot, that added some support for Secure Boot, for reading Bitlocker volumes (in Win7 64-bit!), and that interfered with the BIOS support for Secure Boot on some Asus motherboards. Might investigate that. (Edit: Pretty sure that issue was Asus-specific, and didn't extend to Gigabyte boards, which yours appears to be.)

 

[Flayed]

Maybe try resetting the CMOS with a jumper or something

 

[Mantrid-Drone]

That is the only thing I can think of doing too.

There is a possibility the CMOS battery has died as the MB, whilst 'new' is over 5 years old. It was installed 3 years ago but apart from an early test boot into the BIOS/UEFI the PC was not used and a OS installed until last week.

Would a dead CMOS battery produce the effects described? Seems unlikely to me but I've never taken the battery out of any PC. The nearest I've had is a laptop where the permanent unreplaceable CMOS battery simply died through age (13 years old). The laptop just reverted to factory BIOS.

The odd thing here is that there is no audio or LED fault indication - if it was the RAM or a essential component failure, like the GPU, most MBs, I thought, used some sort of code to indicate the nature of the failure. There is nothing like that with this.

I know it could not have been as a result of overheating either. I'd put in system monitoring software, the system fans were/are all working perfectly and the core temperatures were shown as under 40C minutes before the fault appeared. The fans are also definitely not at full speed which suggested the the default 75% PWM setting must still be in effect.

If this was some catastrophic MB failure which occurred merely by chance after the failed Windows updates wouldn't the fans spin up to full speed without any settings to control them?

Anyone with some idea what might have happened please help.

 

[VirtualLarry]

The odd thing here is that there is no audio or LED fault indication - if it was the RAM or a essential component failure, like the GPU, most MBs, I thought, used some sort of code to indicate the nature of the failure. There is nothing like that with this.

Generally, you need to connect a "beeper" (peizio speaker) to the 4-pin SPKR header among the front-panel headers, to hear audio beeps.

One way to determine if the CPU and mobo are more-or-less working, is to install a speaker, and REMOVE the RAM, and you should get some beeps complaining about "No RAM found", or "Low RAM missing".

If the board doesn't beep, and is simply dead, that's a bad sign, at that point, and probably either bad PSU, bad mobo, or bad CPU. (or missing power cables).

 

[DAPUNISHER]

Weird. Start with a basic troubleshoot. Take it out of the case and boot bare bones using i5's iGPU. Reseat everything, one stick of ram at a time, different PSU, no storage or USB device of any kind connected. You are just looking for a post right now. Once you have eliminated all the usual suspects, then you can feel confident proclaiming the mobo went belly up. I have read about boards sitting a long time failing shortly after getting powered, but I have never experienced it personally.

 

[Mantrid-Drone]

Thanks, both ^^. The GA-Z77-DS3H does have MB speaker built in and I actually added a MB speaker to my other desktop, the one I'm using now, because its Gigabyte MB doesn't. A few pence from eBay to get that comforting POST beep and in cases like this helpful warning beeps too.

I tried clearing the CMOS value first and even taking out and testing the CR2032 battery. Battery fine, just less than 3v, but it did not make any difference.

I had not read VirtualLarry's post but I did do exactly as suggested - removed all the RAM modules but first I tested them in the 4 slots separately. All good. So I then thought why not see what happens when none are fitted. When I heard those warning beeps on reboot it suggested, maybe the MB is OK.

That is apparently correct because it is now booting again. The PSU was not the culprit.

It took disconnecting everything - not just the power and SATA connectors from the SSD, HDDs, DD and GPU which is what I tried to begin with. No, I had to take the GPU, sound card and a legacy landline/answerphone card out of their PCI slots too. Only when I did that and rebooted: POST beep and VGA graphics output prompting me to load optimised defaults.

I still do not know which device it is yet - I'm going to have test each one but my money is on the GPU. But I hope I'm wrong because the other two I can well do without.

Possibly useful info for others who might find this thread later: I had some trouble after booting because I'd forgotten those MB optimised defaults include setting the SATA mode back to IDE from AHCI. The Windows Repair Tool which ran on reboot and might have been useful did not like this either. Whilst the MB's BIOS/UEFI could be navigated OK as normal in IDE mode neither the keyboard or mouse USB connections were powered under the Windows Repair Tool so you could not select any options.

Anyway once that was sorted I booted to my desktop with big ugly 800x600 VGA graphics and used a pre-Windows update restore point. On restart I did a System File Check and have scheduled ChkDsk too.

Annoying thing after all this is that MS has managed to sneak an upgrade to Win10 pop-up onto the PC. Once I've identified the problem PCI device getting rid of that will be a priority.

I'll post again just to confirm which device it was with the problem - if it is the GPU is there anything I can do? Could it be a driver issue or perhaps a botched firmware update AMD attempted rather than a 'simple' hardware failure?

EDIT

It was not the GPU thankfully. The culprit was actually the Xenta sound card. Tried it in both spare PCI slots with the same dramatic result. So now removed and the system getting back to normal.

Unlikely you'll find that sound card on sale anywhere except eBay so even if it looks like a good way of adding TOSLink digital output for little cost my experience suggests it is not and and is best avoided.