Computer won't keep time correctly

[BondiBlue]

I'm running into an annoying issue with one of my computers. It's an older computer (specs below), but it does everything I need it to do without issue. For the last few days I've been having trouble with the clock slowly getting out of sync. The clock seems to be running slower than it should be. I manually synchronized the time in Windows Settings, then I used a separate device to compare the time. After 5 minutes on the separate device the problematic computer had fallen behind by over 7 seconds. That doesn't sound like a lot, but it adds up fast. I left the computer for roughly 30 minutes, and when I came back it was well over a minute behind real time. 

 

Does anyone have any ideas what could be causing this issue? This machine has never done this before, and I have no clue why it would just randomly start. I don't think the CMOS battery is dead, but even if it was would that still cause this issue when the machine is plugged into mains power? 

 

Specs: 

• Motherboard: ASUS P8Z68-V/GEN3 (yes, this machine is old)
• CPU: Intel Core i7-2600 
• RAM: 32GB DDR3-1600
• PSU: Corsair RM750
• GPU: GTX 1080 + RTX 2070 Super
• Drives: 1TB SATA SSD + 3TB Seagate HDD
• BIOS: 0301 (I think this is the release BIOS for this board)
• OS: Windows 10 Pro for Workstations 22H2
• This machine is always connected to a wired ethernet connection

[Ottoman420]

I think the CMOS battery does keep time while the system is off.

 

Not 100% on this so il google

"

The CMOS battery is not there to provide power to the computer when it is in operation, it's there to maintain a small amount of power to the CMOS when the computer is powered off and unplugged. The primary function of this is to keep the clock running even when the computer is turned off. Without the CMOS battery, every time you turned on the computer you would need to reset the clock.

On older systems the CMOS battery also provided the small amount of charge required to maintain the nonvolatile BIOS memory, which remembered BIOS settings between reboots. On modern systems this information is typically stored in flash memory and does not require a charge to be maintained.

--nhinkle"

 

I would change the CMOS battery see if he issue persists but your BIOS will need to be reconfigured after

 

What else could it be unless you are near a nuclear reactor?

 

[RokinAmerica]

Symptomatic of a dead CMOS battery, as Ottoman stated.

[BondiBlue]

3 minutes ago, RokinAmerica said:

Symptomatic of a dead CMOS battery, as Ottoman stated.

Even when the machine is on? Never experienced that on any other machine with a dead CMOS battery. 

[Alex Atkin UK]

4 minutes ago, BondiBlue said:

Even when the machine is on? Never experienced that on any other machine with a dead CMOS battery. 

That would be odd.  With a bad CMOS battery usually everything is fine even when the PC is off - as long as the PSU is on providing standby power.

 

You should even be able to change the CMOS battery in this state without losing the settings, though not necessarily advised as if you drop the battery on the motherboard it could short something out that still has power.

 

But motherboard designs can vary, maybe a quirk with this specific model?

[BondiBlue]

1 minute ago, Alex Atkin UK said:

That would be odd.  With a bad CMOS battery usually everything is fine even when the PC is off - as long as the PSU is on providing standby power.

 

You should even be able to change the CMOS battery in this state without losing the settings, though not necessarily advised as if you drop the battery on the motherboard it could short something out that still has power.

 

But motherboard designs can vary, maybe a quirk with this specific model?

That's always been my understanding as well. This machine runs 24/7 and has done for years, so there is a decent chance the battery is dead. I guess I'll swap it out and see if the behavior changes.

 

In the past I've run machines that are completely missing the CMOS battery and they didn't have this sort of issue. This motherboard does have other strange quirks (mostly to do with the network controller and memory settings), so it wouldn't surprise me if this is just a board thing. 

[Sawa Takahashi]

15 minutes ago, BondiBlue said:

Does anyone have any ideas what could be causing this issue?

The CMOS battery isn't use when the computer has power, for sure. You can replace the CMOS battery since it costs so little, but I doubt it will fix your issue. 

Here are a few causes for RTC (real time clock) drift :

• General drift error
• Ambient temperature drift error
• Aging drift error
• ...

And that is not very helpful in finding a fix....  I think the aging drift error may be your issue.

 

In Windows, you can always enable the auto date and time feature (in settings) but I think it syncs only once a day. I found this link for Windows 10 that may help you set the auto date and time to update more often : https://answers.microsoft.com/en-us/windows/forum/all/windows-10-automatic-time-sync-interval-minimum/000311be-defc-4076-8185-e30878658183

 

Good luck !

[Alex Atkin UK]

6 minutes ago, BondiBlue said:

That's always been my understanding as well. This machine runs 24/7 and has done for years, so there is a decent chance the battery is dead. I guess I'll swap it out and see if the behavior changes.

 

In the past I've run machines that are completely missing the CMOS battery and they didn't have this sort of issue. This motherboard does have other strange quirks (mostly to do with the network controller and memory settings), so it wouldn't surprise me if this is just a board thing. 

You know what might be curious, try running a different OS and see if time drifts the same?

[BondiBlue]

1 minute ago, Alex Atkin UK said:

You know what might be curious, try running a different OS and see if time drifts the same?

I might just have to give that a try. I also just remembered that I have another one of these exact same boards sitting in storage. It's been a few years since I last used it, but IIRC it has a 2600K installed. Whenever I get a chance I'll throw that board on the bench and see if it exhibits similar behavior with the battery out of this board. 

[Eigenvektor]

1 hour ago, BondiBlue said:

I might just have to give that a try. I also just remembered that I have another one of these exact same boards sitting in storage. It's been a few years since I last used it, but IIRC it has a 2600K installed. Whenever I get a chance I'll throw that board on the bench and see if it exhibits similar behavior with the battery out of this board. 

Might be an idea to install an ntpclient that synchronizes more often, given how much the clock appears to drift.

[BondiBlue]

On 10/8/2024 at 11:38 AM, Alex Atkin UK said:

You know what might be curious, try running a different OS and see if time drifts the same?

Unfortunately, Windows decided to install an update before I got a chance to try that plan. Pretty annoying, especially since this machine is configured to never download updates automatically. However, after it installed the update the time is back in sync perfectly. All the settings have remained the same as well. Not sure what was causing the issue, but it's resolved now.