How to diagnose a spontaneous reboot?

My computer reboots, seemingly completely at random, about once every week to two weeks but has occasionally gone months. It just goes from running fine to the POST with no error messages or anything and doesn't seem to be due to heat or usage as it's happened a couple of times when the computer has booted just a few moments ago and is idling. It's been happening for as long as I've had this computer, almost two years. It's happened in both Vista and Windows7.

I strongly suspect it's a hardware problem. But due to the rareish and random nature of the crashes my normal strategy of just removing hardware until the problem stops isn't really practical. My guess would be Power Supply, Ram, or Motherboard. But I just don't know how to test an issue this random and want to figure out how to confirm which it is before I go replacing things. So is there some software or hardware that can be used to test these sorts of errors? I did run memtest86 for about 8 hours without finding any issues. And the power supply is more than capable of running my system.

It certainly seems like a hardware issue.

The first course of action would be to test your ram. Memtest86+ is a widely used diagnostics tool for RAM. I suggest you leave it running overnight or longer and see if it reports any errors.

If your RAM seems to be alright, you can try running a CPU burn-in tester to see if your processor is alright, if it doesn't produce strange faults.

If that doesn't produce any errors you could try to replace your BIOS battery. I have seen examples in which an empty BIOS battery somehow made a system instable. It's also a cheap solution.

My last guess would be the PSU. Replace it with any decent and recent model and test RAM and CPU again.

A note: if you test RAM, leave as much hardware out as possible.

I won't repeat the guides others have written - you should start by disabling automatic restarts and doing a memory test (forget this though as you already have).

AFter this, Take a look at Bluescreenview, this utility will allow you to see any bluescreens that have appeared in the past and hopefully allow you to track down the fault.

As for the fault itself, the number one cause in older machines is Blown Capacitors, I have attached my little guide:

Blown capacitors -

Blown capacitors can be the cause of many "random" problems which appear to be completly unrelated, very annoying and hard to diagnose.

The top should be almost flat (with slight indents out embossed sections depending on specification... look at the middle one) but you do not want to see any big bumps such as the first one or any leakage as the last one.

Hope this helps.

Turn off automatic restart (the instructions are similar for Vista and 7) and see if the computer is BSOD'ing and get the stop codes. You can also look into the event log under "System" and check for critical errors.

Otherwise, "acquire" a copy of PC Doctor and run some extended diagnostics.

Just to add to the good advice of others above I will add my 2 cents. In my experience the first candidates to check for random reboots and hardware problems are your power supply and cpu cooler. If you have multiple drives & devices and a power supply that is getting on a bit then that would be my first guess for causing the reboot. The other alternative culprit can be heat. If the system is overheating then it may be restarting itself - heat should be easy enough to monitor.

Overall - sounds like a good excuse to update to me - buy a new one!