GTR7 Pro fails to POST after AC power loss

Tullagh

Greetings fellow Beelink fans!

I just got a GTR7 Pro and have been putting it through its paces. So far it's breezed through everything I've thrown at it. It's fast, stable, quiet, and cool. Memory and SSD torture tests passed with flying colors.

However, it has this one strange problem: If it loses AC power, it has a ~75% chance of failing to POST, even if it was completely shut down when it lost AC power. It spins up the fans and turns on the keyboard lights, but no video signal ever reaches the monitor. Shutting it down via a long press of the power button does NOT help, no matter how many times I do it. It'll stay in this state until I cut and restore AC power again. At that point it has a ~25% chance to POST, and if it succeeds, it's perfectly fine again. I can restart it, shut it down, put it to sleep, hibernate it, etc. But if at any point I unplug it or lose AC power, I'm back to the POST lottery.

If any of you have seen this sort of thing before, what's likely to be the problem? I'm guessing the power supply, but I know very little about hardware.

My setup is dead simple:

- GTR7 Pro, all original hardware
- BIOS V21 (upgraded from V15)
- Corsair wired USB mouse and keyboard
- Acer CBL272U 27" 1440p 75Hz monitor
- Original Windows 11 Pro installation


Thanks!

bignay2000

bignay2000 replied at 2023-09-14 12:00
I have a GTR 7 Pro with 96 GB ram upgrade running Proxmox.  This happened to me a few times when I  ...

After several hours trying to get the system to post, I finally gave up and left it unplugged overnight.  Plugged the power in and went to lunch.  The system is back online now.   

Other than this one bug, the system is super awesome.  Hope this gets a bios fix soon!

Tullagh

An update for anyone who might be interested:

I've been using my Pro heavily for about a week now, and this intermittent failure to POST after AC power loss remains the only problem I've encountered. Once my Pro is in a good AC power state, it appears to be 100% stable no matter what I'm doing with it.

To summarize, after regaining AC power, my PC is in one of two states: either it can't POST no matter how many times I try, or it's 100% stable no matter how many times I restart it, shut it down, sleep it, hibernate it, etc.

I'm not a hardware guy, but this FEELS like a power supply issue to me. I ordered a pair of extra power supplies just in case (before I discovered this issue), and I'll do some testing and post another update once they arrive.

Cheers!

BrianofMadCity

When I had this issue on older chipset drivers, I had to unplug all but one HDMI monitor. The USB C monitors and the DP2 monitor had to be disconnected for the boot to work. Ask for an updated chipset driver.

dedecou

This could be a DDR5 training issue, have you tried pressing the clear cmos button for a bit after failing to POST?

Taking note that if your system is actually training DDR5, it could actually take a couple of minutes before it actually POST.

I think if it was a Power Supply issue then you would be very hard pressed to boot at all, much less have a lottery. Maybe do a stress test when you are able to boot, to see if the PSU can supply the power necessary when the system is at full power.

Tullagh

This could be a DDR5 training issue

Oh, interesting! Crucial's page on DDR5 training says it can actually hang the POST, requiring another reboot to resume normal operation. That's similar to what I'm seeing.

However, in my case, a normal reboot (long press on the power button) doesn't seem to do the job. I have to cut and restore AC power, but when I do that, the system is likely to want to redo the training on next start.

If DDR5 training is indeed the issue, the question becomes: Why would training sometimes hang and sometimes complete quickly? And why would my system re-train after AC power loss? Shouldn't it only happen when new hardware is installed?

have you tried pressing the clear cmos button for a bit after failing to POST?

When I first noticed this issue, I thought clearing the CMOS might be the way to restore normal operation, so I tried it, but I only pressed CLR CMOS with AC power off, per Beelink's instructions. After some more testing, I don't think clearing the CMOS was relevant; it's the AC power cycle that got me out of the bad state and back to "the lottery".

Taking note that if your system is actually training DDR5, it could actually take a couple of minutes before it actually POST.

I've let it sit and spin for up to 20 minutes. Any idea what could cause DDR5 training to hang sometimes?

Maybe do a stress test when you are able to boot, to see if the PSU can supply the power necessary when the system is at full power.

I've completed memory and SSD stress tests successfully. That's what's so frustrating. The thing seems rock solid, but every time I unplug it or lose AC power, I'm in for an adventure

GarethRossUK

Tullagh replied at 2023-08-02 12:21
Oh, interesting! Crucial's page on DDR5 training says it can actually hang the POST, requiring anot ...

Why are you loosing AC power so often?

I'd hate to opperate any power reliant computer if outages are frequent.

zyero281

GarethRossUK replied at 2023-08-02 19:15
Why are you loosing AC power so often?

I'd hate to opperate any power reliant computer if outages ...

Does it matter?

dedecou

zyero281 replied at 2023-08-03 00:03
Does it matter?

Not really, though it's not good for the OS(window mostly, linux doesn't have an issue as it has a journaling fs).

Honestly, from what you say, it's really weird. To know more, we would need the POST codes, but I don't think that's going to be possible.

Maybe it's some residual charge from some capacitor or such? If you powercycle it(i.e. press the power button) when it's unplugged does it help? That generally helps with discharging any capacitor.

Try doing it with the magnetic charge plugged and the power supply unplugged from AC, as it could be an issue with that. I have a random board that uses a 12V 3A brick, and the capacitor inside is really enough to light up a LED and supply tiny currents(as the board is off) after it's unplugged(and of course, the board is off at that time), so maybe this could be one of your issues.

Tullagh

Thank you all for responding!

I've had trouble posting to this thread until now (moderation issues?), so I sent PMs to some of you.

Tullagh

To summarize, after regaining AC power, my PC is in one of two states: either it can't POST no matter how many times I try, or it's 100% stable no matter how many times I restart it, shut it down, sleep it, hibernate it, etc.

FINAL UPDATE: A new power supply did NOT resolve my issue.

However, I've noticed that my PC is less likely to get into the bad state if the AC power stays off for a longer period of time. So perhaps it's some kind of capacitor discharge issue.