Topic: New X2 Bios (Read 83 times)

hippo · « **Reply #15 on:** May 03, 2005, 11:30:00 AM »

There is no telling for sure with a clone because the board is a copy of another product and the devices on the board are also counterfeits or seconds or pulls. It's very common for rejected parts to get traded retail over there. Often times parts have special requirements and testing is required.

hippo · « **Reply #16 on:** May 04, 2005, 08:13:00 AM »

Perhaps a cell in the flashrom has gone bad. The cells around it are still okay. You could possibly reprogram the original bios back on setting all cells back to factory. The bad cell is still factory as it hasn't changed. Changing the bad cell at the factory is possible with high voltage programming you cant duplicate in the xbox, this is theory 1.

Theory 2 involves around the fact the cpld may only be decoding 256k and that's 25% of the entire 1mb bios but because the error checking in the software flashing the device is checking after the first of four blocks another 5% appears to have elasped and you are dead as a duck at 30% on either bank. Also it maybe the entire 512k isnt accessable or who knows. If the 25% that's written is properly written (it might be) and the cpld doesnt make the jump to the next 256k page then tying A18 of the flashrom to ground might boot that 256k image laid down first if it's a valid image. If you are testing x2 50xx it's not a valid image so try tying a18 low and flashing a 256k bios via hotswap or whatever.

hippo · « **Reply #17 on:** May 05, 2005, 12:44:00 AM »

Bogus they left you hanging at a17

Chancer · « **Reply #18 on:** May 05, 2005, 04:11:00 AM »

QUOTE

However the 50xx X2 512k bios will not flash at all to the chip.

Have you tried with Eurasia.
Yesterday I was messing flashing X2 5032 with evox and it failed a few times. Flashed using the Eurasia boot disc and a bios disc and no problem straight through.

Chancer · « **Reply #19 on:** May 05, 2005, 08:51:00 AM »

Do all other 512K bios files flash to it ok. ie 512K M8

sampoman · « **Reply #20 on:** May 05, 2005, 07:32:00 PM »

Just flashed a 512K Evox M8_16 bios, worked fine. For some reason it wont flash the X2 50xx bios tho :|

HackDaBox · « **Reply #21 on:** May 06, 2005, 06:17:00 AM »

HackDaBox · « **Reply #22 on:** May 06, 2005, 07:04:00 AM »

oh well at first everyone was having alot of trouble flashing the clones but it seems now that most of the duox2 clones are ok. Just stick to the blue ones to avoid flashing trouble on V1.6

HackDaBox !

hippo · « **Reply #23 on:** May 07, 2005, 07:24:00 AM »

Not flashing on a 1.6 is a problem I went through with my chip. There's ony one reason it doesnt flash on a1.6 and that's bad cpld code and you see other problems too. As far as the flashing goes it's sorta of a mystery. I think when the cpld is on that state it has a problem passing certain bit patterns through or something.

cubistproject · « **Reply #24 on:** May 07, 2005, 06:11:00 PM »

I'll give 'er a go tonight...

hippo · « **Reply #25 on:** May 08, 2005, 06:24:00 AM »

A 512k evox bios is just 2 256k ones joined.

Dra60nSlayer · « **Reply #26 on:** May 10, 2005, 06:19:00 AM »

QUOTE(otherguy @ May 3 2005, 03:44 AM)

i highly doubt you could render your chip useless through a bad flash.

Dra60nSlayer · « **Reply #27 on:** May 10, 2005, 08:11:00 AM »

i had no problem flashing another duo x2 from the same batch when not using my "external LPC bus" so i guess it had to do with too long LAD wires

Xbox used was v1.0

cubistproject · « **Reply #28 on:** May 10, 2005, 08:47:00 AM »

Okay. More of the same over here.

For this test I used an original DuoX2 and 2 clones from 2 different manufacterers.

I used an original DuoX2 with the true AMD flashrom.

One clone appears to use a fake ST M49FL080A Labeled SINGAPORE on the bottom, and lacking the ST logo.

And the other clone uses an original ST M49FL080A sporting the ST logo.

I used Eurasia to see if maybe EvoX was the problem. I first flashed the original chip. It went fine.

I then used the clone with the fake ST. It erashed fine, then froze about 30% of the way through. Both banks were erased/corrupted. I was able to recover both banks fine via hotswap flashing from EvoX.

I tried the higher quality clone w/ the true ST flashroms next (the ones from my shop). It erased, then stopped about 30% of the way through. Again, both banks were erased, but were easily recovered via hotswap flashing.

I'm wondering if the problem is not the CPLD coding, as the clones being used are from totally different manufacterers. I'm wondering if the problem is actually the flashrom itself. Or rather, the way that the software deals with the flashroms. All of the clones contain the ST flashrom (or assumed fake M49FL080As). I would assume the CPLD programming would probably be different on them all. I find it especially strange that it corrupts both banks... You woulnd't think it wouldn't write outside of that 512k bank... I'd like to see an ST transplanted onto a true DuoX2 to see if, indeed, it is the CPLD programming or the flashrom handling...

Just a thought. I don't know shit, tho. It seems super fishy tho, that 2x256k bios would flash but not a 512k bios...

So these chips, thusfar, are deemed unusable for the new X2 bios until an alternate flashing method can be used, or until the manufacterers can fix this. I'm going to email my contacts in China today to see if they can get a remedy happening... until then, I'll try a few more flashing methods ie) sunday flash.

Peace

hippo · « **Reply #29 on:** May 10, 2005, 09:34:00 AM »

Some cloner just needs to work some bugs out is all I'm sure they'll be back with something better for the 080.

Author Topic: New X2 Bios (Read 83 times)