Why did my memory burn up at post?

[Charlweed]

I HAD a socket 478, Asus P4C800-E Deluxe mobo that had served me well for years, first as a game rig, then as a Linux server.
Last month, I decided to upgrade its memory from 1gb , to 3gb.
It had two DIMMs of 400mhz CL2 512k Corasir CMX512- 3200C2PRO. This stuff was top-of-the-line when I bought it. They were installed in the blue alpha slots.

So, I checked the manual, and all the recommended memory is obsolete. No big, I bought two more sticks of 400mhz CL3 Patriot 1gb Patriot PSD1G400. I put the new DIMMS in the empty black beta slots, checked the seating etc, and hit the power button.
I was rewarded with a puff of white smoke from the DIMM slots. The new memory burnt up, starting at the slot, and up one trace of the PCB. At least one DIMM of the old memory was damaged too. And I assume the mobo itself. No Socket 478 mobos were readily available, so basically, I had to buy a new PC.

So, why it it burn up? The manual has no warning except pairing the memory in each slot. The Manual says it supports 1gb DIMMS, though none are on the supported list.
Was it CL2 vs CL3? , Too much memory? Bad DIMMS? Wrong Slots? What?

Charlweed

 

[T_T]

Only two possibilities come to mind, and they're both pretty far fetched, but...

1. The voltage was set too high for those modules and they burned upon power up; or

2. An unbeknownst open circuit in the DIMM slot caused an electrical failure. Though you may have physically installed the RAM correctly and firmly, the actual circuit was still incomplete, thus causing the RAM or DIMM to be overloaded.

Like I said, far fetched, but remotely possible.

 

[Charlweed]

Best answer selected by charlweed.

 

[Charlweed]

Ok, I'm going with the bad DIMM slot theory. I guess it could have been some dust or other debris in the slot. Yet one more thing to be paranoid about in the future.