Are there any problems when shutting down a Mini PC?
Are there any problems when shutting down a Mini PC?
The device has been experiencing shutdowns and screen blanking, especially when running virus scans or under load. The monitor stays on with the power light illuminated, and the fan operates normally. Attempts to manually shut it down by holding the power button failed. The issue seems related to overheating, as temperatures reached up to 89°C during scans, though fans increased speed. Dust accumulation was a factor, improving performance after cleaning. It appears a thermal problem is the primary cause. If this occurs during Windows installation, reinstalling might be necessary, but if it's a heat issue, proceeding could risk further damage.
It would be useful to examine and possibly update the thermal paste between the CPU and heatsink. I've noticed some that have dried, cracked, or were installed incorrectly, and reapplying paste has worked in those cases. For this project, consider Thermal grizzly Kryonaut, Arctic MX-6 or Noctua NT-H2.
I intend to swap the thermal paste. Right now I'm tracking temperatures with HWMonitor and haven't seen any readings that seem excessively high. When I mention MBAM, it's about Malwarebytes Antimalware.
I might have identified what could be the cause of the problem, relating it to his wall outlets. I’m not an electrician, but it seems the outlets he’s using for his computer might lack proper grounding. One of the surge protectors he has includes indicators for grounded and protected circuits. The protected light is lit, but the grounded one isn’t. I checked the wall outlets with a multimeter—ground probe into the bottom hole and hot probe on the plug’s hot side. The short circuit should read 120 volts if it’s correctly grounded. When no devices are connected, both plugs display around 25 volts. If I connect a surge protector to that outlet and test the empty one in the same way, it shows about 68 volts. This pattern repeats when I reverse the plug connections. As I mentioned earlier, I’m not qualified, so I’m puzzled by why both readings are low (25 volts) even with no devices, and why a surge protector connected shows 68 volts. It’s similar to what I observed with hot and neutral plugs, which register around 124 volts.
I’m uncertain if this is the source of the issue affecting his computer, but something feels off. I took his computer home to check, and it behaves normally there. Given the older house, it’s possible that few or no plugs are actually grounded. Many still have the two-prong covers. He plans to have an electrician inspect it.
Another possibility is that he has external speakers drawing power from the same outlet, which stopped working. At full volume, he only heard static. I brought them back home and they functioned correctly.
EDIT: The computer continues to shut down, likely with a blank screen at my location. If the problem stems from the outlet, I realize it may already be damaged.
Consider removing the power strips, as they're likely just inexpensive units with weak internal construction, especially since most of them probably aren't. Connect directly to the wall socket, avoiding any strips, surge protectors, UPS batteries, or extension cords. Check if the issue persists. If it does, relocate the system to a different outlet on another circuit or room entirely. I thought this earlier, but your feedback makes it more urgent.
I brought the pc back home and connected it directly to a grounded outlet. The screen remains blank during virus scans, which is the only time it works properly. I’ve been using it for tasks like backing up files and browsing the web, but nothing else. Memtest+ ran without errors or blanking issues.
I haven’t replaced the thermal paste on the CPU yet—I might have misplaced it. I’ll need to get it before proceeding. If the paste has hardened, it could lead to high CPU temperatures even when idle.
He’s already pressuring me to upgrade. I’m considering a prebuilt desktop but am tempted by a mini PC. In his case, a desktop seems better since he tends to keep it running 24/7 and dislikes waiting. I’m trying to persuade him to shut it down or put it to sleep at night. Either way, he’ll consult electricians to see if the plug can be fixed.
When it operates solely during scans, it clearly isn't related to thermal or electrical issues. Consider removing Malwarebytes and performing a custom scan with Windows Defender. It might not be the best approach to have both installed and active at the same time. Defender alone provides adequate protection. There could be a conflict if both are enabled. Try uninstalling MBAM, restarting, and running a Defender scan just on the C: drive using the custom setting. Check if it still functions. If so, attempt reinstalling MBAM, disabling Defender, and retrying. There must be something affecting it if this is the only occurrence.
And indeed, he should turn it off at night. I've never shut down my main system in 14 years except during hardware installation, working inside the case, cleaning inside, or during electrical storms. Otherwise, it stays on but powers off when I step away for more than five minutes. It's simple to set it to wake only when a key is pressed, avoiding accidental activation from accidental bumps. Still, keeping it running continuously could shorten the lifespan of certain parts.
If this doesn't resolve the issue, consider backing up important data and performing a clean Windows installation. This has happened before due to strange registry changes and unresolved problems.
Hello! You can reach our support team via email at [email protected] for any help you need.
I've just performed a Malwarebytes scan on a 5600X system. The CPU usage rose to 70%, and the temperature climbed to +68°C. After the scan finished, usage dropped to 2% and the temperature fell to +29°C. From this, I believe the scan significantly increases the CPU load and consumes much more power than when idle. If the issue stems from heat, a virus scan could potentially cause the computer to crash.
I agree with @Darkbreeze. The absence of a mains safety earth seems unlikely to be the main problem. Coming from a country where most domestic and office supplies have a required safety earth, I only notice non-earthing outlets when traveling abroad. I’m excluding 2-pole "shaver" sockets which are meant for double isolated circuits.
Examining my laptop power supplies, some have 2-pole (figure of 8 connector) while others use 3-pole (cover leaf connector). ATX PSUs typically come with a 3-pole IEC connector. If a PSU or device only has 2 poles, it should be double isolated, which doesn’t need a mains safety earth. A 3-pole input requires a safety earth to protect against faults by blowing a fuse or tripping a breaker, preventing damage from metal casings touching live parts.
You can run an ATX PSU without a safety earth, but you forfeit the protection and risk serious injury if a major fault occurs and you touch exposed live components. The decision is yours. Consider the small risk of electrocution if a serious fault develops inside an ATX PSU that should be earthed, or hire an electrician to install a mains safety earth in your home. I know what I would do.
N.B. I haven’t verified the shape or design of the Beelink SEI 12, but if any exposed metalwork is present, a faulty PSU could make it dangerous if it’s supposed to have a safety earth.
https://resources.altium.com/p/isolated-...thout-fail
Many computer PSUs include Class Y capacitors between Line and Earth and from Neutral to Earth (see the second circuit). These are built to reduce RFI (Radio Frequency Interference).
https://www.engineering.com/how-to-choos...plication/
If you lack a safety earth and the PSU is connected to mains, the two Class Y capacitors act as a potential divider, making the mains earth connection roughly midway between Line and Neutral. A reading of 68V on the "earth" lead suggests your line voltage could be around 136V—possibly too high, but within a reasonable range. It’s not uncommon for both capacitors to have slightly different values.
The 25V measurement is likely from other PSUs in the house connected to mains, which may also use similar Y-class capacitors, raising the safety earth potential above ground. The leakage currents of these capacitors are measured in mA and should only cause a slight tingling sensation if you lightly touch metal parts without a safety earth.
I sometimes experience a buzzing feeling when handling laptop metal cases with certain third-party double isolated USB-C PSUs. As an electronics design engineer, I attribute this to suboptimal circuit design. I don’t feel the buzz with OEM laptop PSUs.
TLDR
If your computer crashes under heavy load, as others have noted, consider replacing the CPU’s thermal compound.
I should have been more explicit in my earlier posts. Now, when it crashes, it occurs during a custom Windows Defender scan on drive C. Initially, I began with a custom scan using MBAM, which would crash during the MBAM process. However, MBAM is now able to complete successfully. Still, starting a Windows Defender scan first causes it to crash. So far, whenever I’ve been testing it, it has only happened during scans whether it’s the defender or MBAM. But when it happens, it was just browsing the web or handling emails—nothing unusual. It depends on how many tabs were open. Whenever I’ve used it for tasks like backing up files or light web browsing, nothing has occurred. Even during a scan, temperatures never reached that high; they stayed around the mid-70s since I cleaned out the dust. While I acknowledge MBAM might be causing issues, I’ve run both MBAM and defender on my systems for years without any problems. With MBAM, I believe extra protection is better. He isn’t very tech-savvy, and with his grandson using it, that’s another matter altogether.
I’m suspecting a hardware or power delivery issue. When it crashes, I can’t press the power button to shut it down—I have to unplug it. Sometimes, if I quickly reconnect the plug, it restarts on its own. Other times, I need to disconnect the cord from the power brick or something similar.
I plan to conduct a stress test using OCCT and examine the hard drive. I also intend to replace the thermal paste, which I’ve had to order and wait for delivery.
EDIT: I downloaded and ran OCCT. The first time I used it, I selected CPU and set it to steady mode. It ran for at least 30 minutes without any errors or crashes, keeping temperatures just below 70. I then ran another test with CPU and RAM in extreme and steady modes. Within a few minutes, it crashed. This time, I was monitoring it closely; the highest temperature recorded was 77. When the computer restarted and OCCT reopened, it still indicated the crash time as 2:43 and confirmed no errors were detected.