Reply 20 of 22, by Romain
- Rank
- Newbie
Hi everybody,
To be more clear and clean, I make a last post to summarize and finalize this crazy thing - if someone else has the same problem.
First this it's about the m918i V1.2A (Fake cache version), not the m918 (with cache slots avalaible)
So, now I'm pretty happy because the issue is fixed.
The story was :
1- The mainboard worked nicely, but because of DMA conflit, I wanted to move the JP4.. But I moved the JP5 instead .. on 1-2 position (because of my too big hands I suppose..).
2- So, I booted the pc : all was OK
3- Turning off before to sleep
4- The next day, no boot + 8 bips in loop
5- So I moved the JP5 to full open position (as initialy)
6- Ok, boot etc, all ok.
7- The next day, I turned on and … looping bip in loop blank screen
On the internet I found some very old and hard to find topics forum, with some posts that said that sometimes like "this motherboard doesn't support no more its own BIOS anymore" 😜
Now I think it's because of the JP5 that this happens.
The worst thing is that even if you put it back in its original position, the motherboard will keep killing its own BIOS. I managed to reproduce the problem : see attached video below. 😀
It's totally crazy, I've never seen anything like this before 🤣
You can see the attached video the see the bug reproduction, made by the a chip model AT29C010A (but I don't really think that the eeprom is the problem), model also installed on other PcChips motherboard models.
Anyway, I'm happy to be able to reproduce this problem as much as necessary, it shows that this card contains another trap (besides the FAKE Cache), which makes it even more « glorious ».
To solve the problem, the solution was :
- To take a different eeprom reference (here a NM27C010Q-150)
- To replace the BIOS version « 41-C100-001437-00101111-101094-486PCIMB-U » by the « 41-A300-001437-00101111-101094-21011/33-E », easily found.
But because of this update, the motherboard is now unable to recognize the double-sided RAM stick … such life..
Well, finally now it has become very stable - I have tested several reboot, looping benchmarks for several hours with a 486 DX4-100 it works great now.
I spent time for that, but for the curious, I attached below :
- The dump of the original eeprom corrupted.
- The bin file used to flash a new eeprom on a AT29C010A.
- The dump from the AT29C010A after corruption.
- The video of issue reproduction
That can be instructive..
And of course the video with exactly the behavior I had.
It seems to be a very unknown problem, and the only two topics I've found are very old with no help at all. So here it is, I don't know but even if it could be useful to just one other person I'd be happy because I spent a while on this problem which was completely twisted.
Cheers.