What's wrong with this board?

[Spork829]

Hi everyone. I have an Asus H81 board I'm using for my home server/NAS that has a strange issue.

 

Often when I power the system off, it won't start again. Power button does nothing. The only recourse seems to be to shut off the PSU and wait at least a few hours (I usually just leave it overnight) then I can power it back on. Sometimes when it powers off this doesn't happen, and it turns right back on. Seems random.

 

I think it might be a CMOS battery related issue, because when it does start again after the PSU has been off, the BIOS has been fully reset. But I replaced the battery and that didn't change anything.

 

Anyone know what this issue is, or if it's fixable?

 

Thanks!

[Spork829]

I should add, this is a recent development. This board used to be in my main machine and had no issues. This is only happening recently since I started using it again in a new system. Could it be a PSU issue? That's the only relevant piece of hardware that's different than before.

[Guest]

Bios issue perhaps? May not be a battery or psu issue. After all the board is about 7 years old..... 

Tough one to diag without swapping a bunch of parts around. Weed out the usual suspects ya know?

Take a good look at the capacitors on the board. Look for swollen or leakers??

 

[HenrySalayne]

Have you tried to just bridge the contact for the power switch on your mainboard? Your power switch might be broken.

[Spork829]

4 minutes ago, ShrimpBrime said:

Bios issue perhaps? May not be a battery or psu issue. After all the board is about 7 years old..... 

Tough one to diag without swapping a bunch of parts around. Weed out the usual suspects ya know?

Take a good look at the capacitors on the board. Look for swollen or leakers??

 

 

I'll take a look at the caps right now. Unfortunately I don't have too many things to switch out, and I've already tried different CPU and RAM. 

 

What exactly do you mean by bios issue, and how would I diagnose that? Thanks for your help!

[Spork829]

4 minutes ago, HenrySalayne said:

Have you tried to just bridge the contact for the power switch on your mainboard? Your power switch might be broken.

Yep, shorting the pins directly gets the same result.

[Guest]

11 minutes ago, Spork829 said:

 

I'll take a look at the caps right now. Unfortunately I don't have too many things to switch out, and I've already tried different CPU and RAM. 

 

What exactly do you mean by bios issue, and how would I diagnose that? Thanks for your help!

Just like anything, a bios can wear out so to speak.

Ive had boards tell me I have a new cpu after every restart.

My Maximus X which is pretty new will start and stop 3 times. A clear cmos usually repairs that issue, and occurs typically after overclocking a lot. (I use it for benching)

Had another board set v-core to 1.60v when input 1.56v. 

Just a few examples there. Fully working hardware other than some bios quirks. 

[Beerzerker]

It's possible to have a partially corrupt BIOS due to a weak CMOS battery in place. It's not been mentioned if the battery has ever been replaced or not so check that and look for voltage of at least 2.9v's or more.

 

A new battery has 3.0v's so if it's at or below 2.9v's you need to replace it or risk this happening. A weak battery can cause some weird stuff to happen as well, esp upon bootup of the system. 
You can first, replace the battery and then do a reflash of the board to eliminate this as a possible cause. Only other thing related would be a bad BIOS chip and that's something you'd have to replace period or get the board serviced, depending on if the chip is removeable or not.

[Spork829]

27 minutes ago, Beerzerker said:

It's possible to have a partially corrupt BIOS due to a weak CMOS battery in place. It's not been mentioned if the battery has ever been replaced or not so check that and look for voltage of at least 2.9v's or more.

 

A new battery has 3.0v's so if it's at or below 2.9v's you need to replace it or risk this happening. A weak battery can cause some weird stuff to happen as well, esp upon bootup of the system. 
You can first, replace the battery and then do a reflash of the board to eliminate this as a possible cause. Only other thing related would be a bad BIOS chip and that's something you'd have to replace period or get the board serviced, depending on if the chip is removeable or not.

Thanks for the info! I have already replaced the battery, but I will look into reflashing the BIOS, I hadn't considered that a bad battery could corrupt it.

[Beerzerker]

Yes - I'd go for a reflash to make sure all is OK with it.

Once done that hopefully will take care of it. If for some reason the flash fails and it indicates a bad sector/block on the chip then it's bad and nothing you can do except replace it.

There is a guy on fleabay that goes by "BIOS Depot" and you can get a new chip from him, preflashed with the most recent BIOS. You do have to tell him the exact specs of your board and be sure to have all that info available for him.

 

And...... If your chip isn't removable then it's time to send it off for repair.

[aisle9]

What's wrong with this board? A lot of things. For one, I still don't have an answer on why @Fasauceome hasn't been banned yet, and--oh, wait, you mean your motherboard.

 

Which ASUS board is it? I've got some recent experience troubleshooting the H81M-C and H81I-PLUS.

[Spork829]

15 minutes ago, aisle9 said:

What's wrong with this board? A lot of things. For one, I still don't have an answer on why @Fasauceome hasn't been banned yet, and--oh, wait, you mean your motherboard.

 

Which ASUS board is it? I've got some recent experience troubleshooting the H81M-C and H81I-PLUS.

H81M-E, Originally from a prebuilt.

[HenrySalayne]

31 minutes ago, Spork829 said:

I hadn't considered that a bad battery could corrupt it.

It shouldn't. Rewritable flash is basically a charge trap and the cells keep their state without any voltage. Some cells may lose data after some years because the charge slowly degrades. The battery only supplies power for the RTC and to keep the settings.

 

[aisle9]

52 minutes ago, Spork829 said:

H81M-E, Originally from a prebuilt.

The latest and (probably) last BIOS for that board is 3602, from April of 2018. Is that what you're running?

 

That said, I don't think it's your BIOS at all. That's a hardware problem. Take the board apart, check all the connections (especially from the PSU) and blast them out with compressed air. It sounds like you might have a fleck of dirt or a speck of dust in between your PSU's cable and the board connector. It could also just be that your PSU is dying, but I like to run through troubleshooting for other factors before saying, "Just buy a new PSU".

[Spork829]

7 hours ago, aisle9 said:

The latest and (probably) last BIOS for that board is 3602, from April of 2018. Is that what you're running?

 

That said, I don't think it's your BIOS at all. That's a hardware problem. Take the board apart, check all the connections (especially from the PSU) and blast them out with compressed air. It sounds like you might have a fleck of dirt or a speck of dust in between your PSU's cable and the board connector. It could also just be that your PSU is dying, but I like to run through troubleshooting for other factors before saying, "Just buy a new PSU".

I've never updated the BIOS on this thing so I'm sure it's ancient, I should try that. I've tried a couple of different PSUs now so I don't think that's the issue, but I'll check the pins in the connectors on the board, that might be something. Thanks!

[Beerzerker]

17 hours ago, HenrySalayne said:

It shouldn't. Rewritable flash is basically a charge trap and the cells keep their state without any voltage. Some cells may lose data after some years because the charge slowly degrades. The battery only supplies power for the RTC and to keep the settings.

 

I know "It Shoudn't" but with Mr. Murphy on the prowl he can make it so - Older DFI boards for example can and will act crazy with a weak battery.

Boards such as a Socket 939 Ultra D exhibit this behaviour when a battery gets weak in it and it's capable of corrupting it's BIOS, other DFI's of different models do the same.

An Abit AN7 is the same way because it's a battery eater thanks to the onboard Uguru chip. When it's battery gets weak it too acts out and replacing it takes care of it. Most of the time it's fine with the BIOS itself but I've seen those become corrupted before as well, just not as often as the DFI's did.

 

I am using older boards as examples here but the basics still apply.

[HenrySalayne]

@Beerzerker
How do you determine if the BIOS is damaged because the battery is old and weak or because the flash cells are old and the charge degraded? Does it mean we should store mainboards without the battery? Does the controller wake up from time to time and refresh the flash cells, so a good battery will prevent charge deterioration?

You might be right, but a weak battery and a degraded charges in flash cells are both signs of age and there might be no link between these two phenomenons. I certainly don't know. Would be interesting though.

[Beerzerker]

19 hours ago, HenrySalayne said:

@Beerzerker
How do you determine if the BIOS is damaged because the battery is old and weak or because the flash cells are old and the charge degraded? Does it mean we should store mainboards without the battery? Does the controller wake up from time to time and refresh the flash cells, so a good battery will prevent charge deterioration?

You might be right, but a weak battery and a degraded charges in flash cells are both signs of age and there might be no link between these two phenomenons. I certainly don't know. Would be interesting though.

The biggest indicator is buggy behavoir period, happens alot when you have a system that's setup properly in the BIOS with all else being right but just refuses to boot, even at full default settings or it just does some stuff that doesn't make any sense.

 

I don't fully understand it all myself but from what I've ran into before is from where I speak.

 

The system does read and could write a tad of info to and from the BIOS chip everytime it boots up but as to what, it depends on the individual system.

 

We do know whenever you install and activate Win 10 it writes a tad of info that's OS activation related to the BIOS chip, tying your system to the OS copy. If you do a hard reset by removing the battery to clear it you'll still be able to boot it up without it throwing the "Windows needs activation" message or other screwy shiite so that's not battery-dependent in itself.

 

An example of a more battery based retention of the OS activation is once I actually had to fix a machine for a friend using Win 7 and it was demanding activation of the OS everytime the person booted the system.

 

Activating windows took care of it until the next time the system was booted, once shut down it would repeat this problem at the next startup.

In this case the OS activation was held by the battery instead of being fully written to the BIOS chip and would get cleared out everytime the system was shutoff, requiring it to be reactivated everytime the system was ran.

The machine itself required no reconfiguration of things in the BIOS itself setup related but I checked the battery and it was around 2.8v's or so, too weak to function properly.

 

After I replaced the battery in the system it stopped doing that and retained activation of the OS.

 

So.....

With some systems it does make a difference, with others it doesn't but never hurts to suggest checking and replacing it should it be weak - It's there for a reason.

 

As for storing a board I would remove the battery period and worry about it if and when I went to use it again. Probrably won't hurt it to leave it in for awhile (Not long term as in for several years or more) but that's what I do with the ones I have stored in the closet.