My V3 bolt has shutdown on me twice since I have had it. The system is on a UPS, and we have not lost power recently. The first time it shutdown on me, I was not around. This I was, and the machine was just sitting idle. The temps had been running between 38c-44c. Of course, the fan was running. After the shutdown, I unplugged it, and felt the heat sink. It did not feel hot. It was warm at best. When I tried to power it back on, it did not start up right away. It took several tries. When it booted, the fan did not turn on right away. In fact, it turned on and off a few times, then just stayed on. The temps were less than 40c Has any one experienced this?
What kind of power supply? USB-C or Jack? What kind of OS? Did you see any events in the eventlogging of the OS?
Sorry for the lack of detail. My OS is Fedora 30. I did not see any entries related to event. That is, nothing that would lead me to believe that the OS decided to shut the system down. My power supply is plugged into the jack. 19V 3.5A. That is plugged into a UPS so power fluctuations should not be an issue.
Does it shutdown properly (should be visible in /var/log files) or does it crash? Crashing could still be power source related. Is it possible to use another adapter/wires? Did you install extra drivers? Did you install/update the AMD display drivers?
I don't believe the OS is the issue. It does not appear to be crashing. It seems as if the power is being turned off. I did check the power settings to make sure that it is not suspending. I did update and install AMD specific drivers that I could find. Fedora is good with AMD. I am going to test my UPS. I have no reason to believe that it is not working correctly, but that could be my problem. My son has a Mac book pro, so I will give his power supply a try. I believe it fits the specs. Since I don't know what is going on I am not leaving the machine on when I am not using it.
re: power supply is plugged into the jack. 19V 3.5A Showing exactly make/model would be helpful. Some extremely cheap one may have one or more issues: - quality parts and design meeting the spec. - insufficient output side (DC side) filter capacitor <-- I have seen this many time
Okay, it looks that a non-OEM type of power supply 19V 3.42A (DC cable seems to be very long and not so thick, it has a thermal shutdown function, BTW.) At this point, there is no easy way to check if this power supply is well behaved all the time. (I have proper equipment that's why I know in a different SBC with a similar type of power supply, it did not have a large enough output capacitor so voltage dip when demand ran high.) I can only suggest get one of the "tested working" power supply in this forum and try again.
I will check the temps on the power supply when the machine is on. I did not consider that it was getting too hot given that the machine has shutdown when not in use but on for more than a 24hrs. I think this is probably my problem. My plan is to test with my son's Apple power supply. Although, I might just purchase a known good power supply. I would rather use the power jack over the USB-C port. Thank you.
I still have not had the change to try a different power supply, but it did shutdown on me again. It had only been on for about 2hrs. It was essentially idle most of the time. It was idle when it shutdown, but I was there. The CPU temps were less than 45c. The power brick was barely warm, so I don't think a power supply thermal shutdown happened. I checked the logs and did not see any evidence of the system doing a proper shutdown. It just powered off. I will test with a different power brick, but the Arduino still never turned off. Its light never stopped flashing. Also, I had problems getting it to power up after not being used for 4 days. I press the power button and the lights would go green for a second or two, then flip back to yellow. After 4 tries, it powered on. This is a switch that is connected to the front panel header. It is also the first time this has happened.
Well, it has turned off on me twice today. The first time after about 2.5hrs after I turned it on. The second time it had been on for a while, with a shell open and xsensors running. I stepped away for about 5 minutes and came back and it had turned off. This time it had been on for about 4 or 5 hrs and the temps had gotten as high as 60c, but it had been around 40c before it turned off. I don't think it is the power supply as it is not getting hot. It is not even warm. I am wondering about the cpu cooling. I am going to change the fan and use a Noctua fan. I might replace the thermal paste when I replace the fan.
Well, it just turned off for the third time today. This time I had Brave open with a shell and xsensors. The temp was 37c.
I would install the OS again, perhaps something went wrong there. Also check the OS logging again for clues.
I now think the issue is my power supply. I started to investigate getting another one to try. In reading the reviews of many of these, I noticed that several reviews said that the power supply would just turn off for a little bit. Not a huge problem if you are charging a laptop, but it is a big problem if you are running a desktop from it. I suspect my power supply is just turning off briefly. Again, would not be a problem for a laptop. I don't think it is a thermal shutdown as the power supply is barely warm. I don't think it is the OS because the system just turns off and there is nothing in the logs to indicate that a shutdown was initiated. I did replace my fan with a Noctua fan. The Noctua fan is running off 5v from the Arduino buss. The fan header on the board did not provide enough power to run the fan. The fan is quiet, but will now run as long as the device is getting power. The upside is that the thermals are low. I did not replace the thermal paste. Thermals are staying around 39c with a browser open with several tabs. The room I am in is 77f.
See my response on #6, specifically - insufficient output side (DC side) filter capacitor <-- I have seen this many time Those SBCs (Intel or AMD) use lot of CPU bursting. This creates quite a lot of instantaneous power demand. Thin DC power cable, insufficient DC output capacitor will cause the power to drop suddenly. Although it can be very brief, (view this as a sub-millisecond brown out), the CPU simply can not work under such condition.
