Cold Boot Maximus Formula

softnerd pc perspective said:

I've had an interesting development with this. I was about to pack this thing up and try to get rid of it, and just before I put it in the box I tried hooking it up on my test bench one last time - sans motherboard tray - just the naked board.

It POSTed fine. I was puzzled, so I left it all day to get cold while I was at work...still connected to the PSU to simulate the power situation. When I got home I powered it up and...it posted. What changed? Well, I had a chasis fan plugged into the PWR FAN header that was now unplugged (kept forgetting to unplug it in testing because it's part of the mobo tray), and it was no longer mounted to the tray. I already confirmed that it wasn't grounding out against anything on the tray, so I tried plugging the chasis fan back into the PWR FAN header. Bam - hung at COUNTMEM. Unplugged it, and it POSTs again. I was able to reproduce this until the capacitors got warm and it was POSTing every time.

So, I left it overnight and tried again this morning. First without the fan - it POSTs, then with the fan - hangs at COUNTMEM. Interesting.

I have a hunch that this board has some wonky design quirks when it comes to the +12v1 power tap. Power supply fans usually get power from the PSU - not the header, but they need a tiny bit of power supplied for the fan monitor/speed control to work. If you look at the trace routes from the PWR FAN connector the lead directly to the EATX CPU power plug. Now, I got started thinking about this when people said they solved this issue by plugging in the 8-pin plug instead of the 4-pin, or realized that the plug wasn't pushed in all the way.

Is it possible that a full-on chassis fan is pulling enough power away from the +12v1 tap that there isn't enough juice getting to other components when the capacitors are cold? So far I've been able to reproduce the issue 3 times. I've re-assembled my entire machine. If it POSTs properly when I get home tonight I'd say there's a good chance I found the culprit.

To sum up:
- Use the 8-pin EATX CPU power plug.
- Make sure it's fully connected.
- See what happens if you avoid plugging anything into the PWR FAN header. Weird...I know, but it so far it seems to make a difference for me.

Not sure if this is going to solve the problem, but I'm hopeful.

Click to expand...