Cold boot fine, warm restart hangs and STOPS the CPU FAN with Q Code 1F

[somalezu]

A couple of days ago I built a new system using a new C7H WiFi. This is used as a workstation with a couple of Virtual Machines, so no overclocking was applied (only DOCP standard to get the DDR4 to the nominal speed).
The system is stable but I noticed a big problem restarting (at least 4 or 5 times already).
When restarting from a clean installation of Windows, after the operating system is finishing shutting down and the system should reboot, the monitor gets dark, the system hangs and the CPU FAN STOPS.
The Q code shown on the C7H diagnostic display is 1F.
Waiting for up to a minute will not change anything - the system seems stuck and will not try to reboot, only the CPU cooler metal sink gets pretty warm (because of the stoped CPU FAN).
Turning off the power (by long pressing the power button on the case) and turning on power once again will boot the system normaly.
This system should run 24/7 but now I'm afraid of Windows updates and automatic nightly restarts (to apply windows updates). If the system will hang with the CPU FAN STOPPED during an automatic update, I'm afraid the CPU would get cooked.

Configuration
Ryzen 3900X on a C7H Wifi, 2703 bios. stock cooler connected to the CPU_FAN plug on the MB, PWM mode / standard profile, spins and reports fine in AI Suite 3
RAM: Corsair Vengeance 3200Mhz 2x16 GB (CMK32GX4M2B3200C16).
Video: Asus GT 1030 SL BRK
Samsung 970 Pro 1 Tb Nvme

[UncleHunk]

Did you try with older bios 2501? There are people who are affected by restart hang bug, but fan error is something new.

[somalezu]

Haven't tried an older bios version yet because I am still strugling to replicate the problem in a controlled manner.
Not every warm reboot fails - I rebooted for almost 10 times in a row today without any problem. Is it cured? I doubt it, because I did not do anything special.
it randomly failed to reboot about 5 times in the last three days or so, every time with the same error code (1F) and with the CPU FAN not spinning.
It's hard to identify a solution when the problem seems to pop randomly.
I'm running out of ideeas.

[DrHappyAngry]

I've actually had this issue on the CS VI, but didn't note the Q code when it happened and haven't rebooted much the last week. I've only used firmware 7403 with the 3900x. One one reboot from Ubuntu 18.04 It got stuck booting and I had to hit the power button. The only time I've rebooted it in windows, it got completely stuck and I had to reset the cmos just to get it to bootup again. There's no overclock, just DOCP standards to get the ram set 3200mhz with the right timings. Firmware 7201 didn't have the issue with the 2700x. This is almost exactly like the crap the original firmware when the CS VI was first released.

[Ryo_z]

yup, this reboot issue is what i reported earlier in another thread.

But, it seems like asus is ignore it and someone even tried to downplays the issue just because they don't reboot / warm boot the pc that often and also the only thing they care about is performance, not stability / any other thing.

https://rog.asus.com/forum/showthread.php?112279-X370-X470-AGESA-1003AB-Bioses&p=782466#post782466

[DrHappyAngry]

Reikoji mentioned in that thread that it only happens if the PSU is turned off, and mine definitely was left on for both reboots that got stuck. I only eventually tried turning off the psu power when it wouldn't boot up at all, but it didn't help and I just wound up resetting the cmos.

[AndreiR]

I am actually struggling with the same issue with my new C7H + 3800x + 64GB of ram (DOCP). It seems to work fine with 32GB, but as soon as I put 64GB it happens again. I've downgraded to 2501 bios and the issue seems to have disappeared. In general I would get random hangs on boot with newest bios, but 2501 seems to be stable.

[somalezu]

After my initial post the system became pretty stable, I tried 30 restarts in a row with different bios settings (with or without DOCP, with or without core boost, energy saving, and so on), memory sticks connected to the "FlashBack USB port", WiFi turned on and off ... and everything looked just fine, no more hang ups when restarting from windows.

Today I updated the BIOS to version 2801 and now the problem is back and is very easy to reproduce.
If the DRAM timming is set to Auto, Aida64 reports DRAM at 2120 Mhz and Windows restarts without a glitch.
If I activate the DOCP standard profile in BIOS (witout any other change from the default settings), the Windows seems to run just fine at DRAM nominal speed (about 3200 or just a bit lower) but on every Windows restart the system hangs with QCode 1F on the onboard motherboard display and CPU FAN stopped.
Same thing happens if I manualy set the DRAM profile to 3200 Mhz: windows seems to work very well, but if I attempt a restart, after the system shuts down all devices, the motherboard tries to restart, enumerate a bunch of q codes on the embeded display for a second until 1F is shown and hangs with the CPU FAN stopped (CPU FAN RGB is still on and cycling from color to color, PSU seems to be on, CPU heat sink gets warm).
Before reverting to previous BIOS version, I want to enable DOCP / SPD timmings once again and run a stability test (should I try some memory test for a long period of time? the stability test in Aida64 ? do you have any sugestion for a relevant memory test ?
Do you have any suggestion of any BIOS setting I should test to check for an influence on the Windows restart issue?

The issue is pretty important to me: this computer should run 24/7 and I am afraid any Windows/software update that attempts an automatic restart will hang the system WITH THE CPU FAN STOPPED. Not sure if at that point the CPU is still able to throttle when reaching some temperature limit.
So at this momment I have to choose between running DRAM at 2300 Mhz or risking some overnight restart that will cook the CPU...

Ryzen 3900X on a C7H Wifi, 2703 bios. stock cooler connected to the CPU_FAN plug on the MB, PWM mode / standard profile, spins and reports fine in AI Suite 3
RAM: Corsair Vengeance 3200Mhz 2x16 GB (CMK32GX4M2B3200C16).
Video: Asus GT 1030 SL BRK
Samsung 970 Pro 1 Tb Nvme
PSU: Seasonic Focus+ Gold 750W

[somalezu]

With BIOS version 2901 the problem seems to be resolved (Loaded optimised defaults then I set the memory profile to DOCP / SPD 3200 Mhz and I had no issues until now, after 3 or 4 restarts).
I will come back and report if I will notice any restarts issues in the future.
Thank you,
Cosmin

[Heavy_Chevy]

I've had the same issue with all bios versions with the same memory you have the only thing that seems to help the most it's the up the memory voltage to 1.4 all other settings left stock... I've had the same random reboot failures as you different types of memory any settings the only thing that seems to help the most is to up the memory voltage .05 above recommended settings.... I even have two sets of the 32gig kits you have for 64 gigs and I can run/reboot all day long no problems as long as I up the voltage 1.4....