Gigabyte GA-870A-UD3 won't power up sometimes

eldrin

n00b
Joined
Jul 25, 2013
Messages
33
Hi All,

Lately I've been having some trouble with my system based on a Gigabyte GA-870A-UD3 motherboard (Rev 3.0).

Sometimes when I push the power button the computer doesn't start. No fans, no beeps, etc. The only thing that happens is that the PSU (Corsair AX850) starts making a faint buzzing sound.

What I then need to do is switch off the PSU itself using the button on the back. And then after switching it back on, try turning the PC on again. Sometimes I have to do this several times.

Today after trying this unsuccessfully a few times, I eventually unplugged the 2 power cables (ATX and ATX_12V) from the motherboard and then plugged them back in. Then the PC was able to start successfully.

I've got a Corsair AX850 in my other (Intel-based) desktop PC as well, which has never given me any trouble, so I'm inclined to think the motherboard is at fault rather than the PSU.

The problematic PC is using a Radeon 6450 and I read a forum thread elsewhere where someone solved the problem by switching to an nVidia card. Do you think it's likely that the GPU could be the cause, or is it more likely the motherboard?

Does anyone have any suggestions?
 
Sounds to me like you may have a short in the psu, Do you have another unit you can try, or use the one from your intel setup? try to unplug the power switch from your case to mobo and just use a screw driver to short the connection to start the rig up to make sure its not a bad power switch.
 
Thanks for the suggestions Hellbent88. I've eliminated the power switch on the case as a culprit. And now that I think of it, I'm sure I encountered a similar problem when using the PSU that was installed in this system before upgrading to the AX850. So I strongly suspect the motherboard is at fault. Unfortunately, after getting the system powered up and shut down, it typically has to be left for quite a while (at least overnight, but often longer) before the problem can be encountered again, so I can't quickly do experiments to track it down.

One thing I disabled quite recently in the BIOS was ErP, which initially seemed like it solved the problem. The USB ports now remain powered when the system is off, and it looked like this made a positive difference. But then a day or two later, the system wouldn't turn on again, so that isn't it.

I wonder if there's any possibility that the GPU could be involved. The Radeon 6450 is powered from the PCIe slot only, and maybe the power it requires at startup could be a factor. I suppose I could purchase a spare GPU that uses a PCIe power cable to see if that makes a difference. But I'd be interested to hear whether the GPU theory is even plausible.
 
Last edited:
Well, I've just swapped the AX850 PSUs around, i.e. taken the one from the perfectly working Intel system and put it into the problematic AMD system, and vice versa.

So after a few days I should have a definitive answer as to whether the PSU is to blame.
 
after using a number of gigabyte boards over the years i've come to love the cold boot problems that sometimes follow.

Dual bios seems to eat up motherboard batteries a little faster than you'd expect. Might try putting a new one in.
 
Thanks Aaron_ATX, that's a good idea for something else I can try.

However, I always thought that the symptom of a dead motherboard battery would be the time getting lost, or the BIOS settings getting lost, and none of those things are happening.

Is it really possible that a dead battery could prevent a cold boot?
 
Seen gigabyte boards freeze in various ways with a dead mobo battery. All before POST.

If your board is relatively recent it probably has a dual bios. If so a dead (or more specifically dead shorted mobo battery) can keep it in a state of limbo while cold booting as it cannot access either bios for a POST.

If that doesnt do the trick start pulling power connections to things one at a time while off and see if you can isolate the offending device.

5 bucks on the mobo battery tho :)
 
Thanks Aaron_ATX, that makes sense.

To give an update on the other experiment I'm currently trying with the PSUs switched around between the 2 systems: Today the GA-870A-UD3 cold started successfully the first time I tried (which was after being off for about 38 hours). And the other system is doing fine so far too, although I haven't left it turned off for an extended period of time yet. I wish there was a way to test more quickly than switching off for 1.5 - 2.0 days before trying to switch on again.

In the meantime I'll also look at ordering a replacement battery so that if necessary I can do that test too.
 
Oh forgot to mention. You shouldnt have to leave the powersupply unplugged for hours on end to discharge the caps.

Turn the machine off, unplug it, switch the PSU on/off switch to off, then hold down your power button on the case for 10 or so seconds. That should bleed off any juice in the capacitors.
 
Hey eldrin, still keeping an eye on this thread. Good call Arron on the bios battery, never thought of that. don't hafe to order one, hell you can get them almost anywhere. You running the newest bios for your board? might help. when you go into bios does it list your power stats 12v. 3.3? have had older gigabyte mobos but they only said shit like ok with no stats. are your temps good on your cpu and not making your rig shutdown and not start again? When you said you thought it might be video card causing the prob, you should be able to remove it and boot your rig and just get a no video card error but it should still boot and cpu fan should run and case fans too.

Just Food For Thought :D
 
Hey Hellbent88, yes, I'm running an up-to-date BIOS and all the stats concerning voltages, etc. are fine. When the system does start, it's completely stable.

To clarify for all, the problem manifests itself in the following way:

1. I do a full, graceful shutdown of the operating system.
2. The power cable is left plugged in to the PSU and the power switch on the PSU is left in the on position.
3. Even though the operating system is fully shut down, the motherboard still has trickle/standby power to it -- this can be seen, for example, by the fact that the USB ports are still powered.
4. After many hours (approx 1.5 - 2.0 days), if I then press the power button on the case, all that happens is that the PSU starts making a faint buzzing sound.

The last point is why I was saying I need to wait so long between tests. The problem scenario is starting the system from a state where the motherboard does have power from the PSU, but the system has been in a shut down state for a couple of days.

Update on my current experiment: Just moments ago I've started the Intel system which now has the AX850 that was previously in the AMD system. This was after having been shut down for just over a day and a half. It started first time. So the summary of the first round of my experiment is that after swapping the AX850 PSUs around, both systems have started successfully. Hmmm. I'll need to repeat this a few more times over the coming days...
 
Update: Since swapping the PSUs, both systems continue to power up.

However, the GA-870A-UD3 is now displaying more frequently some different problems I'd encountered sporadically in the past. Often, the system will just freeze part-way through the POST, usually just after enumerating connected drives. Other times it'll freeze just as Windows 7 starts loading. From reading various forums, these problems seem to be quite widespread with this board. And getting into the BIOS is tricky. Despite pressing Del many times quickly, sometimes it just continues to boot anyway, and other times it recognizes that Del was pressed and says 'Entering Setup...' but then just freezes.

In the end I've decided this motherboard has just been too much trouble to deal with, and it's wasted too much of my time. I've just gone ahead and ordered an Asus Sabertooth 990FX R2.0 to replace it.
 
In the end I've decided this motherboard has just been too much trouble to deal with, and it's wasted too much of my time. I've just gone ahead and ordered an Asus Sabertooth 990FX R2.0 to replace it.

Should be real happy with the Sabertooth, buddy of mine has one and loves it.
 
Will do. It hasn't been delivered yet, and when it is it might take a few days before I have a chance to install it.
 
I've now installed the Asus Sabertooth 990FX R2.0.

So far so good. I haven't encountered any hardware problems yet, and it seems to be running in a stable manner. There are some minor things that I can note at this early stage:

1. The audio chipset (Realtek ALC892) is inferior to what was on the Gigabyte board. With the Sabertooth being much newer, more expensive, and with it being positioned as a high-end board, I was surprised by this. Most annoyingly, while all the analog audio jacks on the Gigabyte board supported re-tasking, only a single rear audio jack (front audio out) on the Sabertooth supports this. Previously I had one jack configured as 'Front (speakers)' and another as 'Headphones (speakers)', but now because only this one jack supports re-tasking I have to choose Front or Headphones, not both. Researching this online, there appears to be a registry hack that some people have used as a workaround, so I'll try that out later.

2. Using the Gigabyte motherboard, I used to get a WEI score of 4.4 (for Graphics) for the Radeon 6450 installed in this system. Using the Sabertooth, I get a score of 4.3 for that same test. Weird. The scores for the other components are the same.

3. When installing the drivers for the 3 onboard ASMedia USB 3.0 controllers, I had to do 2 sets of installations for each of the 3 controllers in Device Manager. For each one, after pointing Device Manager to the driver, it would install the 'ASMedia XHCI 1.0 Controller' device and then immediately find another Unknown Device. Then I would need to update the driver for the unknown device and point it to the same driver folder again, so it could then recognize the associated 'USB Root Hub'. This was the first time I'd encountered having to install drivers in this fashion, in 2 identical steps per device.

4. I like the EUFI BIOS. This is actually my first experience working with a UEFI system and I find it to be a very pleasant way of doing things. The only thing that has me a bit perplexed so far is that when it comes to boot devices, somehow the system designates a single device as 'UEFI' for booting from and it's not clear to me why it won't/can't apply to multiple devices simultaneously. For example, if I have no USB flash drive attached, one of the selectable boot devices is 'UEFI: <name of optical drive>', however when a USB flash drive is attached, then I see 'EUFI: <name of flash drive>'. So does this mean if I want to do a EUFI mode boot of a CD then I need to make sure there's no flash drive inserted in any of the USB ports at startup? It's very confusing.

5. I used Paragon P2P to adjust the OS installation to new hardware before booting Windows for the first time on the new motherboard. It worked quite well and afterwards I only had to manually install drivers for those components that were different to what I had before (e.g. the ASMedia USB and SATA controllers). I had to re-activate Windows 7 Ultimate afterwards by phone because the automatic re-activation failed.

And those are my findings/thoughts so far. Next, I'll play with tweaking settings, customizing fan speeds, etc.
 
Some further updates:

The WEI score for Graphics actually changes between 4.3 and 4.4 depending on whether I have my desktop displayed on just a single monitor, or extended to a second monitor.

The USB drivers have a setup program that appears to install all needed components at once -- it wasn't necessary for me to install the drivers manually using Device Manager.

The boot options update dynamically based on what devices/media are available/inserted when the system is started. It turns out if I have a bootable CD and a bootable flash drive inserted, then I see EUFI boot options for both. Booting from a flash drive has to be explicitly chosen as a boot override though - when it's just put in the boot priority list it gets skipped over.

The Sabertooth set my memory timings as 11-11-11-28 even though the SPD timings are actually 9-9-9-24 for the speed I'm running the memory at. So I had to manually set these timings in the BIOS.
 
Back
Top