PC Powers On Then Off Then On

Associate
Joined
28 Jun 2017
Posts
89
So sometimes the PC works fine, and sometimes not.

The issue I’m having is I can game on it for a few hours. Then turn it off at night. When I turn it off I let it turn fully off, then turn it off at the power supply and plug and then turn off at the wall and press the power to discharge it fully. I then turn it back on a few days later and sometimes it’ll power up right away no issues and sometimes turns on, then off, then on.

It did throw up a Kernel Power Error 41 so I ran the tests and it came up with two issues and fixed them via the /scannow.

This issue fixed it for at least 4-5 days and went to turn it on now and didn’t get a CRITICAL Kernel Power Error but threw up a bunch of errors in Event Viewer but only Errors and Warnings like Hyper-V-Hyper, Kernel-Boot, Device Setup Manager, LSA etc.

Just ran a RAM test and that came back all clear.

New GPU, Motherboard & Power Supply was replaced recently.

Anything anyone can suggest as it’s driving me crazy trying to figure it out!

Ryzen 3700x, ASUS B550, RTX 4070, Corsair Dominator 2x8GB 3600Mhz Ryzen Tuned, Alseye AIO cooler and Super Flower 750 PSU
 
Last edited:
When I turn it off I let it turn fully off, then turn it off at the power supply and plug and then turn off at the wall and press the power to discharge it fully. I then turn it back on a few days later and sometimes it’ll power up right away no issues and sometimes turns on, then off, then on.
Why do you discharge it fully? There's no need to do that, or to use the power supply switch. I'd just turn it off at the wall.

sometimes it’ll power up right away no issues and sometimes turns on, then off, then on
Can you describe this in detail? I suspect what's happening is that the board thinks something went wrong and is doing a reset, but you could only confirm that by taking my suggestion above ^^

but threw up a bunch of errors in Event Viewer but only Errors and Warnings like Hyper-V-Hyper, Kernel-Boot, Device Setup Manager, LSA etc.
A lot of errors and warnings in event viewer are nonsense and can be safely ignored.
 
So sometimes the PC works fine, and sometimes not.

The issue I’m having is I can game on it for a few hours. Then turn it off at night. When I turn it off I let it turn fully off, then turn it off at the power supply and plug and then turn off at the wall and press the power to discharge it fully. I then turn it back on a few days later and sometimes it’ll power up right away no issues and sometimes turns on, then off, then on.

It did throw up a Kernel Power Error 41 so I ran the tests and it came up with two issues and fixed them via the /scannow.

This issue fixed it for at least 4-5 days and went to turn it on now and didn’t get a CRITICAL Kernel Power Error but threw up a bunch of errors in Event Viewer but only Errors and Warnings like Hyper-V-Hyper, Kernel-Boot, Device Setup Manager, LSA etc.

Just ran a RAM test and that came back all clear.

New GPU, Motherboard & Power Supply was replaced recently.

Anything anyone can suggest as it’s driving me crazy trying to figure it out!

Ryzen 3700x, ASUS B550, RTX 4070, Corsair Dominator 2x8GB 3600Mhz Ryzen Tuned, Alseye AIO cooler and Super Flower 750 PSU
I sound like a broken record at this mount but have you updated the BIOS?

I know I say it in every other thread but it really is a good way of ironing out bugs (unless you get a bad updated BIOS which does happen).

It sounds like the system isn’t happy with one of the components and AMD is usually a bit more hung up on RAM than Intel so it might be training the memory over and over again but hard to say.

Is your memory running at full speed or does it default to JDEC speeds? 2666 or whatever.

CPU-z in Windows will tell you.

Also, what memory test did you run because the version you should be running is Testmem5 with the Anta777 extreme preset.
 
Why do you discharge it fully? There's no need to do that, or to use the power supply switch. I'd just turn it off at the wall.


Can you describe this in detail? I suspect what's happening is that the board thinks something went wrong and is doing a reset, but you could only confirm that by taking my suggestion above ^^


A lot of errors and warnings in event viewer are nonsense and can be safely ignored.
I guess it’s a habit I got into when not using to out of the risk of any static…I’ll just hit the switch and plug socket plug from now on.

So in detail I’ll hit the power button, computer goes to post and then turns off immediately, then back on with no issues. But the initial power on then off of course is what I’m concerned about!

I’m going to try avoid fully discharging it to see if it fixes the issue.
 
I sound like a broken record at this mount but have you updated the BIOS?

I know I say it in every other thread but it really is a good way of ironing out bugs (unless you get a bad updated BIOS which does happen).

It sounds like the system isn’t happy with one of the components and AMD is usually a bit more hung up on RAM than Intel so it might be training the memory over and over again but hard to say.

Is your memory running at full speed or does it default to JDEC speeds? 2666 or whatever.

CPU-z in Windows will tell you.

Also, what memory test did you run because the version you should be running is Testmem5 with the Anta777 extreme preset.
BIOS has been updated. I like to keep up to date with it!

I set up the profile to run the memory at full speed @ 3600Mhz

Weirdly I’ve never had the issue while playing a game or anything usually on Idle or on boot up mainly!

I used the In-Built Windows Memory Diagnostics
 
So in detail I’ll hit the power button, computer goes to post and then turns off immediately, then back on with no issues. But the initial power on then off of course is what I’m concerned about!
When this happens, does the board clear the BIOS settings, or just do a double boot with no other ill-effects?
 
I’ll have to have a look where to find this it’s a ASUS PRIME B550
APM Configuration
- Restore AC Power Loss

In some circumstances, boards with this feature can power up and go "Oh, power, Hi!" and then they think about if power was on or off and if it was previously thought to be on (e.g. there was assumed to be a power cut), then they will boot up again to restore whatever was the last powered state.

Now that you're not draining the power, this might not be an issue anyway.
 
APM Configuration
- Restore AC Power Loss

In some circumstances, boards with this feature can power up and go "Oh, power, Hi!" and then they think about if power was on or off and if it was previously thought to be on (e.g. there was assumed to be a power cut), then they will boot up again to restore whatever was the last powered state.

Now that you're not draining the power, this might not be an issue anyway.
Will have a look into the BIOS menu later on when I’ve got a minute, just hoping it’s something simple!
 
Will have a look into the BIOS menu later on when I’ve got a minute, just hoping it’s something simple!
For what it is worth, I had an old Intel motherboard (might still have it in a box here somewhere) and for that generation of chipset a double boot was considered normal behaviour :o

I think they did finally fix it with a BIOS update at some point.
 
Here is a description of DDR4 memory training in detail https://www.systemverilog.io/design/ddr4-initialization-and-calibration/

Found this thread here on things to change for an MSI BIOS, it may help because all the manufacturers have similar features


Memory training often goes through a reboot cycle multiple times to self test parameters, try new parameters and retry to make sure that the system is stable.

Manually setting variables forces the system to POST with set parameters and if those parameters are off, it causes the system to fail to POST.

AMD systems often take AAAAGES to POST. Like 5 mins. This often is interpreted as a faulty system, especially on first power on.
 
Back
Top Bottom