My PC has a strange cold boot problem.

mrstabhappy

Honorable
Jul 27, 2013
13
0
10,510
Hello, I have this really strange problem that I can't seem to diagnose on my windows 7 pc. It seems that after an certain amount of time turned off the next time it cold boots, It refuses to load windows drivers at a certain point. It passes BIOS every time but it seems to hang at classpnp.sys and ends with a bluescreen error of 0x000007B.

A hard reset fixes this (I Physically use the back PSU switch to turn it off) and even then it might not work (It may take several times) once every driver loads the PC is fine and can boot property every single time. It seems that cold boots are the source of the problem but even then it has to be a certain amount of time that passes after shutdown ie. 8+ hours based on my observations.

My PC specs
CPU I7 4770K
GTX 770
750 watt PSU
24 Gigs of Ram
1 SSD
2 HDDs + 2 external HDDs
A PCIE USB 3.0 expansion card
DVD Drive.
 
Solution
Bummer, sounds like a thermal contraction and expansion problem. I would start by moving the cables to the booting hard drive. Try new cables or unplug them and plug them into a different port. You will be lucky if that is the cause. Often it is a crack in a circuit trace or a leg of a chip that pops its solder from the solder pad. It makes a hard to find crack that opens when the chip cools. Then it closes 8 or 10 seconds after the chip get power and heats up. If you have any older hardware in the machine that was made using the first generation of lead free solder. The solder was pretty brittle and would pop from the traces. 2009 ? I don't remember the date just that I had a bunch of electronics fail early be cause of the...
Bummer, sounds like a thermal contraction and expansion problem. I would start by moving the cables to the booting hard drive. Try new cables or unplug them and plug them into a different port. You will be lucky if that is the cause. Often it is a crack in a circuit trace or a leg of a chip that pops its solder from the solder pad. It makes a hard to find crack that opens when the chip cools. Then it closes 8 or 10 seconds after the chip get power and heats up. If you have any older hardware in the machine that was made using the first generation of lead free solder. The solder was pretty brittle and would pop from the traces. 2009 ? I don't remember the date just that I had a bunch of electronics fail early be cause of the new solder.

Anyway, when your machine is cold, you can use a hair dryer and heat up a component before you power up to see what might have a problem. For example, heat up the drive and try to boot. Next time it is cool, heat up the hard drive, do this until you figure out the part. Most people just give up and start replacing parts. Last time I did this it took 8 tries over weeks and I found the defect under a heat sink and I confirmed the Break with a optical stereoscope. You can also use a can of compressed air to cool components also to cause the failure.
 
Solution
Wow it looks a lot more complicated that I thought it would be, so the problem might lie on the boot drive right? switching cables does nothing. I was thinking about upgrading the ssd to a higher capacity model anyways.
 
problem it can be anywhere in the system, it could be the drive, it could be chips on the motherboard
if you are lucky it will be in something you can replace.

I had a machine that would do this every morning, I would just power it on, wait 10 seconds and warm boot it with control+alt+delete right after it booted. Some days I would forget and a hour or two later it would bugcheck and I would have lost a bunch of work. It made me mad and I spent a unreasonable amount of time and effort to figure out the cause and I had to see if for myself. a bit of solder holding a chip leg to the board had come free. It was a address line to a memory chip. if it was connected the memory location changed by what ever offset the leg controlled. The effect was in the morning when the chip was cold, windows would load a driver's data over that memory location, the chip would heat up and when the electronics tried to access the memory it would point to a new location and the drivers data would not be there and the system would bugcheck. I could see the data in memory and it looked ok but it was offset by a block that was controlled by that one memory address line., .