My Windows PC has been randomly shutting down, sometimes multiples times a day, for no obvious reason. It’s not the RAM, it’s not the heat or fans, it’s not any particular software, it’s just dying I guess. To be fair, though it has a lot of nearly new peripherals, the CPU and Motherboard are now quite crusty and old, so it’s probably just age.
Any suggestions on model or brand or whatever? Looks like the PC landscape has changed since my day. I need it for 3D and video editing, and also a little bit of gaming, but most of my peripherals (monitors, hard drives, video card etc) will be ported over. I guess I can’t just start it up by using my current C drive running Win10, but everything else should work by just plugging them in.
Note I am in Australia, so certain name brands may not be local to me.
If your machine is otherwise adequate to your needs, you might consider buying it a new power supply and seeing if that fixes the problem. Especially if some of these newer peripherals are powered by the PC rather than their own independent power supply. You might find a US$30 or 50 power supply solves your problems. If not, it’s not such a large wasted expenditure that it crimps your purchase of a whole new desktop unit.
Also, have you investigated the Windows Event Log to see if there are any clues about what the problem is? It sounds like youve done some of that if you’re sure it’s not RAM or temperature, but …
If you’ve added any expansion cards to it since new, not just recently, try removing those one at a time and see if that alters the failures. If, e.g., your GPU card is going wonky that can crash the whole show. Triggering an unplanned shutdown is a bigger stretch than a mere crash, but ACPI has a number of mysterious byways to force sudden shut down.
It’s a big kerfuffle to unplug important things like GPUs. I don’t think it’s that, anyway, as…
…the Event Log shows nothing consistent. For a while I thought it was the Windows Update doing something, but I repaired that file and now it shows nothing at all, so I think that was unrelated coincidence. I think whatever it is, a new PC will solve it all, and in any case, my MB is over ten years old (I think I got it in 2011) so even if it’s not responsible, it’s def on its last legs.
If Motherboard is from 2011, it probably is time to get a new PC.
My oldest computer in use is 2014 and that was demoted to light use at the gaming table and not every day use.
I’ve had this problem with used computers I bought online that were around that vintage. In both I wound up trying replacing the power supply and the motherboard, and the problem still eventually came back. In fact, the second computer was a full replacement for the first.
So I would agree that it makes sense to get a new computer. What gives me pause is your reusing your peripherals, as it’s at possible they could be part of the problem–especially the GPU.
What specific GPU is it? I assume it’s probably newer than your motherboard, if you use it for 3D, video editing, and some gaming. But, still, it’s possible that you could replace it with something cheap enough to avoid having to worry about it.
Similarly, SSDs are getting quite cheap. And a 2011 mobo won’t have NVMe, so you’d probably want at least to replace the boot drive. If you want a slower device for pure storage, you could also buy an SD card (and a SD-to-USB dongle if your new computer doesn’t have an SD slot). If you have some very large hard drive that you want to use for extra storage, you could always attach that externally with USB adapter (or case), which would avoid most potential for problems. That said, a bad HDD is not as likely to be the cause of the shutdowns.
In short, I’d like more info about your system. Do you have an SSD already? What GPU do you have? What CPU are you running, and how much memory?
I know you probably want a computer that is better than the one you had, but this will at least let us set a minimum of what you should get.
I had this happen to me a few years ago. It was the power supply (which surprised me since it was a good one and more than adequate to power my system).
If you are thinking you might build a new system I would suggest buying a new power supply (with an eye to future power requirements). Swap-out the old PSU with the new one in your system and see how that goes. If you still have problems you can bu the rest of a new PC. If it works you saved a lot of money and hassle (unless you just want a new PC anyway…totally get that).
Also, buy a good power supply from a reputable manufacturer. It seems an un-sexy part people/companies like to go cheap on. Don’t. Pay a bit more for a good one (ask here for recommendations if you want that help).
Anytime a PC randomly shuts down, especially if it was working fine for some time and then shuts down without any prior error messages – then the cause is almost always overheating.
So you (or a qualified tech) should:
1- open the cover and blow out several years worth of dust.
2 - check that the fans are clean, lubricated, working, and spin freely.
3 - check that the main chips (CPU, GPU) are properly pasted to their heatsink.
4 - check the power supply for overheating, capacitors, etc.
But note that hiring somebody to do this might cost nearly as much as the resale value of an old computer (especially if it needs any parts replaced). So it might be time to consider buying a new one, while this one is still working and can be used to transfer things over to a new one. Then put a note on it about the problem, and donate it to a nearby electronic charity center.
I have done 1 and 2 already. Not tested 3 or 4. However, it seems like my PC does not restart randomly after I take the side cover off and let air through. So overheating may very well be at play here. And if it’s the PSU, then maybe I should do as @Whack-a-Mole suggests and just buy one of those as an initial step.
So please, which are reputable brands of Power Supplies? My current one is an Antec.
That is not meant to be a definitive list but it is a good start. And, of course, you need to sort out your power needs and features which may change which PSU is best for you. Most of that will be your CPU and your GPU. Add in case fans and/or water pumps and that is most of it usually. In the old days (like three years ago) 750W tended to be plenty. With newer, power hungry CPU/GPUs 1000W or more may be in order.
As an aside…watch for the 12VHPWR connector which is unique to the most power hungry Nvidia graphic cards (also a little notorious for melting).
ETA: Bigger power PSUs are…bigger in size. This may be an issue for your case. Most will fit but the 1000W and bigger ones can be an issue in some cases. Something to keep in mind.
A more likely issue is to make sure your current computer has a standard 24 pin connector on the motherboard since some prebuilt office systems by Dell, HP, etc have proprietary motherboards and connectors. Not sure on the origin of this particular system.
Though, for a 2nd gen system, I’d be thinking hard about a full replacement anyway.
600W PSU would probably suffice for that (400W GPU and 105W CPU max power plus fans/pumps). I’d go for 750W though.
It is worth noting that while a PSU may say it is good for (say) 600W it is not good to run them at 600W. They can manage a power spike there but are not meant to be run there. Kind of like running your car engine at max RPM. It can do it but you don’t want to live there. Some headroom is in order.
Also, consider if you will build a new PC in the future. Newer CPUs/GPUs are power hungry. Much moreso than what you currently have.
Aim for “silver” or “gold” in your next PSU. It is an efficiency rating which also suggests a level of quality (I’d go for gold unless you need to save a buck).
You can get “platinum” or “titanium” but those are much more expensive and I do not think worth the money except, maybe, in some edge cases.
If you’ve already done 1 & 2, and improved air circulation like running it with the case open … then it’s is even more likely to be overheating.
Check #3 for dried out/worn out thermal paste. ($8-$25 DIY fix).
If that doesn’t do it, then the power supply is next most likely. But before buying one, see if you can check it out with a borrowed one. Ask to borrow a compatible one from a friend or computer geek. Or many local computer repair shops will let you test it out with one of their power supplies, if you agree to buy the replacement from them.
I’ve had a few laptops and desktops that I suspected of having thermal problems. In each case, I set up a fan or a hair dryer (not on a heat setting) to ventilate them temporarily. (You know…remove a side panel and aim the fan inside, or set the laptop on the edges of a couple books and direct the hair dryer at the bottom vents.) About half the time this cured the problem immediately. I would then go on to clean the interior, replace a fan, and reinstall the thermal paste.
One laptop had a dead fan, but a replacement was only about $14. Replacement was pretty easy. Until I found a new fan, I ran one of those cheap USB cooler pads under it and that by itself temporarily cured the problem.