I have done this -- It doesn't cause the problem. I had the machine off for two days with the PS unplugged (no aux power), and the power connected disconnected from the MLB. Booted just fine after that. I think whoever mentioned the CAS setting being too fast might be on the right track, but it's weird that the factory BIOS fixes it (for a while).
WRT DMA issue:
(I only booted using LB)
Also, last night, I _hammered_ the southbridge by copying 50M files on the HD, FTPing 50Meg files, and Copying 100M files from usb stick to HDA. I ran this for several hours. The CPU heatsink got pretty warm. No crashes. Previously, just copying from HDA to HDC would crash it, or from USB >> HDA, regularly.
So I Power down, Reboot (LB again), restarted the HDA >> HDA copy and the FTP. Froze up within 10 minutes.
I'll re-run the tests tonight using the factory BIOS -- who knows what's going to happen.
Ronald G Minnich wrote:
Eric Poulsen wrote:
I'm conviced this is the case -- I have too many weird issues that can be fixed by simply flipping back to the factory BIOS, turning the system on, getting a "bad CMOS" error, then immediately powering off and switching to LB, which suddenly works again.
You could try this: do the exact same test, but turn off the board, *disconnect the power connector*, let it sit 30 seconds, and then put in LB and see if it works or not.
I'd be interested to hear this result.
ron