The computer occasionally fails to post or start.
The computer occasionally fails to post or start.
I built a new pc with all brand new parts (except spare HD) and I just cant seem to figure this out. The system is comprised of the following:
Ryzen 1300x stock wraith cooler
Windows 10Pro x64
Asrock X370 Killer SLI/AC ATX
Adata Spectrix D40 8gbx2 Pc2400 DDR4
Evga GTX 1060 6GB
EVGA 550w Bronze
Rosewill case with 4 fans
USB 3.0/2.0 header plugged into the board (2 separate connections)
Internal PC speaker (added for troubleshooting)
Sata3 2TB Hitachi 7200rpm spindle drive main drive
Sata3 500mb 5400rpm spindle drive old drive with old stuff on it
When i turn on my PC, sometimes it will not post. The system comes on, fans are running, memory lights up, mobo aura lights up but no post and no video. There are no bios beeps when the issue happens. I have left the PC for 20 minutes and it never posts. When it will not post, occasionally the reset will work but usually i have to hold down the power button. Some times it takes 1 try some times up to 4, but the faster the clock speed the more often it happens and requires more multiple reboot attempts before post. I also have noticed that its slower to post the faster I go on the processor OC. At first I thought maybe I just lost the lottery and got a junk processor but Im not so sure now as this even happens at what I would consider normal range for the processor. Im starting to suspect the MOBO. If I set the clock at 3.8, it doesnt want to boot/reboot 3 out of 5 times, if set it at 3.7 its once out of every 10 and even happens at 3.6. Ive never left it clocked at stock clock long enough to know if it does it at 3.5 but i suspect it will. At 3.7Ghgz I dont feel like am getting crazy with my speeds. I currently have it @ 1.26875 volts nothing else tweaked CPU wise and I have the memory at 2933 MHZ at 1.35v using the XMP 2400 profile (16-16-16-39). The issue happens even when clocked at 2400 and stock voltage. When the system does post, I can get into windows, play games, run stress tests, browse, whatever... Ive gone days without rebooting and playing multiple 4-5 hour sessions. Its completely stable and never gets above 78 when stress testing at 3.8. 3.9 gets up into the 80s after a few minutes of the testing so I shut it down although it plays games fine on 3.9 and stays in the low 70s. Ive tested using Intel burn test, Cinebench15 and Prime95 I am not getting memory errors or cpu errors or crashes at any speed 3.8 or slower when running stress tests over night so I do not believe it is unstable. Here are the things I have tried:
Slowly stepping higher voltage on the CPU up to 1.395 at both 3.8 and 3.9, 3.8 is temp and math stable at 1.29875v
Using one stick of ram at a time
Using A1+B1, A2+B2(recommended), A1+A2 and B1+B2 memory slot configurations
Disabled Multi-threading (SMT) because the Ryzen 3 doesnt support
Disabled AM4(?) memory test on the mobo (recommended in a thread on Asrocks site)
Set XMP for memory verified the issue happens at all memory clock speeds
Upped the Vcore +120mv
Unplugged the power and pressed power multiple times to drain caps (still will not post most times)
Cleared Cmos with jumper and battery
Upgraded to Bios 3.0, 3.1, 3.2 and 3.4
Re-seated all Power/cards/memory (did not do the processor yet, need thermal paste)
Disconnected external devices and case usb connections from headers for testing
I did also check on a few forms looking for issues particular to Asrock and did come across a couple posts but nothing like what I am experiencing. Most relevant info I found on Asrocks page where others occasionally report the no post issue. Not sure if this is just still an issue with Ryzen overclocking and the bios not being up to par or if this is a hardware issue. Next week I will be trying a totally different brand of ram from a friends PC to see if its any different. I feel like I have thoroughly tested different voltages up to the recommended maximum with the only real difference being CPU temps going up and I have eliminated it being 1 stick of ram. I have a hard time believing the slight OC on my processor is too much but maybe I am wrong. I feel like the ram is a shot in the dark but worth it to try and also thought I would check here before I spend 270$ on a new Asus MOBO. Any help is appreciated. Thanks in advance.
I'm experiencing a problem with my current PC. Usually, I need to switch off the power strip completely after turning off the computer. If I just turn it off through Windows and leave it, it won't work after an hour or so. Trying this might help. I'm not entirely sure what's causing it, but it could be related to the PSU. If that doesn't resolve it, you can try holding the power button on the PC for about 30 seconds while it's unplugged. This should discharge the capacitors, which I've used before fixing similar issues.
I've attempted to hold the power and press the power button repeatedly for 30 seconds while the device is unplugged. Often, it doesn't respond.
It has been the same problem no matter what BIOS was used. This is why I updated the BIOS.
Notchback:
It's been the same regardless of BIOS updates. This is why I changed the BIOS.
I'm not sure what exactly the problem is, but it could be many things. Maybe I should try using someone else's PSU to see if that helps. If it works, I'll return it for repair and get a new one.
I tried that but I'm also going to wait for some other suggestions. I don't see how overclocking would make things worse, though it's possible because of the extra power usage from OC. When I check the hardware monitor, the power looks within the expected range and all voltages are normal. EVGA PSUs are known for their reliability from what I've seen online, so switching to a Seasonic wouldn't seem necessary unless my PSU is failing. If there are many discussions about dirty power or high failure rates, that would change things.
I expect to receive another power source on Monday, along with additional memory, and I plan to evaluate both and share my results. Do anyone have any suggestions?
I tested HWMonitor during my gaming session at 3.8 tonight and noticed the vCore voltage on my CPU dropping to as low as .648 volts. That doesn’t seem normal. I plan to apply some thermal paste tomorrow to re-seat the processor. What do experts say about this voltage dropping so low? Usually it’s around 1.29 to 1.31 according to the software.