imsolidstate

6970m power issues

I was recently working on getting an AMD 6970m working in an R1 M17x. I didn’t get it working, but that’s probably largely due to the fact that I was testing with a bad card.

When the computer powers on, before it POSTs, it applies voltage to the three power rails for the graphics card. The graphics card is supposed to turn on its regulators and when things stabilize assert the MXM power_good signal. This happens before any signals are applied to the card’s busses. However if one of the regulators is out of spec, the power_good signal remains low and the computer will not even start POST so you won’t get any error codes, just a black screen.

I looked up the datasheets for the card’s two main regulators and measured the voltages produced. The regulator that I think probably runs the memory chips (APL5913) was spot-on but the regulator for the core (ISL6228) was off by about a tenth of a volt. Hence the failure to release power_good and continue boot. However my card was underclocked when I got it, so the voltage was commanded at .9V but was actually at 1.04V. Since none of the resistors or capacitors on the board are marked and I don’t have a schematic, it’s not really fixable. But since I knew 1.04V wouldn’t hurt the GPU core, I decided to just cut the power_good trace and see if it would boot. It was dead anyway so I didn’t have anything to lose. You can do this easily by cutting the trace from pin 6 of the MXM connector, plus there is enough trace there to jumper it back together with some solder. Turns out it worked.

With the trace cut the open-drains were removed for the signal line and it was allowed to be pulled high. The computer started up just fine and the video works. Because I know that the voltages are running high, I can just go in and manually drop all of the voltage settings in the vBIOS. It’s not as good as fixing it properly but it beats throwing it in the bin.

I did go back and test my R1 motherboard with the system BIOS that I modified the MXM information on when I got the card working. I had a hunch that the power and thermal definitions in the MXM block were preventing the card from POSTing. So I modified the structures for power and thermal and fixed the checksum. It didn’t work though, there must be something else in the BIOS holding it up. Since I got an R2 motherboard running in my machine, I’m not particularly motivated to tear my machine down and keep trying so I will probably abandon getting the 6970 working on an R1 board.

Tags: Alienware, Electronics