Can't run X3 BIOS?

[Reaper450128]

I've got a problem that seems similar to a lot of others on here but nothing has resolved it so far. Here's what I've got:

I originally started with a v1.6b Xbox and the X3 chip a few years back. I used the latest v1.6 X3 BIOS on it with no problems for years. A few months ago the mobo died and I recently got an eBay used Xbox to replace it. The replacement is a v1.0. I pulled all the necessary parts from the v1.6b box and transplanted the chip into the v1.0 box (which worked perfectly unmodded) using the pin header install process (desoldering sucks but my wire header was long gone) and got it running last night.

I knew I needed to put a 1.0-1.5 BIOS on the X3 so I flashed it with the latest non-v1.6 X3 BIOS. Flashed fine from the backup ROM with FlashBIOS 3.0.1, but upon starting it with the newly flashed BIOS all I get is a brief power on, no video other than a quick flash, then it turns off. The LEDs on the X3 are turning on, the LED on the X3 front panel switch is blue, the ring on the eject button is green for a second before it turns off, then the whole box turns off about 3 seconds later.

So, to try to rule things out I flashed the latest 1MB Cromwell BIOS, and it loads fine. Flashed a 256K EvoX M8+ and it works fine. Used EvoTool to make a 1MB BIOS from the EvoX one and it works fine. The only BIOS that won't work it seems is the X3. I've checked and retouched all my solder joints on the pin header at least four times, all have continuity at about 0.4ohms. Any ideas? I can use the EvoX BIOS but I'd gotten to like the X3 one, so I'd like to get it working.

I should note that when I accidentally flashed the 256K EvoX BIOS with the switch set for banks 1234 it did something similar but didn't stay powered as long. Switching to a single 256K bank fixed that, as did using the resized 1MB EvoX BIOS. I also tested 1MB BIOSes in both 1MB banks with the same results; Cromwell works in either X3 works in neither.

 

[Martin C]

Your install is showing all symptoms of a dry joint on pin 13 or 14 - the ones which go to the EEPROM.

The fact that it's only the X3 BIOS which uses these pins further supports this, given the fact that every other non-X3 BIOS has worked for you.

Definitely re-check these pins and ensure you have continuity.

 

[Reaper450128]

Blah... I don't wanna take it apart again! I've done it so many times now I can see where every single torx fastener is in my head without even trying, lol

Ok, I'll check them out now. Just as a test are there points on the board for those two pins I could use my meter on to check continuity with the corresponding points on the back of the X3? That way I can be sure I've got them all good before I put it back together for the nth time.

 

[Reaper450128]

OK, found what I needed, I think. Based on Unicron's HDD password recovery tutorial I think I see the points on the EEPROM pins 13 and 14 go to. There is no continuity from pin 13 to EEPROM pin 6 (the SCL pin) so I think that's my problem. I cannot get the joint to work so I suspect I'm not getting solder through the board to the ring on top where the trace that runs there is. I ran a wire directly from pin 13 to EEPROM pin 6, sort of a wire rebuild on a single LPC pin. Before I fry something, am I thinking right?

 

[Reaper450128]

OK, I'm impatient so I went ahead and tried it with my fix, and sure enough all is well again! Not sure why I can't get that pin solder properly so it can work via the board trace but with it wired to the proper leg on the EEPROM X3 3294 loads up great and I have ConfigLive back. Thanks for the help!

 

[Martin C]

Glad you fixed it!