PCI FIREWIRE PROBLEM WITH 8500/200 - any help? (OrangeMicro FireWire PCI)

[IDinnitDuit <brave@new.world>]

Sorry 'bout the crosspost, but it's relavent...

Do any of you who happen across this message also happen to have a
PowerMac 8500 (specifically an 8500 - not an 8100 or 8600, etc, but
MAYBE a 7500) that's successfully using some sort of Firewire
interface? I know it's a shot in the dark, but I'd like to know what
kind of card you're using for the FireWire ports. The only [supposedly]
compatible card I could find locally was the OrangeMicro FireWire PCI
Board, but apparently my machine is the hidden asterisk in the
compatibility list. So before I take a big crapshoot and try to guess
which of the cards I'll have to order online would be the best
alternative involving the least amount of return-postage, I figured I'd
see if anyone here might have had better luck with any other particular
Firewire stuff on an 8500.

The only FireWire device I'm using is a 7200RPM IDE drive in a case
made by MadLogix, but I had the exact same problem with a
fresh-off-o'-Fry's shelf Maxtor IDE/Firewire rig before that. Not to
mention that I had 99.999% success when using either drive on my
B&W/300. Both drives were bought brand new by me and yes I have already
done all of the stuff I could/should do with extensions, PRAM, magic
fairy dust, etc. 
 
Well... that is... all i should do except perhaps for that one thing
I'm not thinking of. 


Oh, and if you're wondering about what's actually wrong, it's that when
I try to copy something, even relatively small, from the FW drive to
the 8500's internal SCSI (two Fast Wide Fujitsus striped together) the
data gets corrupted somewhere along the bus. If the file's fairly
large, the copy will fail with an unknown disk error. If it's small, it
might complete the copy sans errors, but the new data will be unusable.
Also, when I try to launch an app from the drive, well... take your
pick. Sometimes it hangs hard, sometimes I can Cmd-Pwr or force quit
out, sometimes I'll get far enough to see a program's splash screen
blow technicolor chunks all over itself. Note that as often as those
things occured, the data still on the Firewire drive was always tip-top
as it flowed over the G3's Firewire bus

Actually, when I cut the drive's throughput in half, it seemed to
perform a slight bit better (faster) on the G3, and a bit worse
(quicker discos, more erratic processor stalling, etc) on the 8500.
hmm... digital negative feedback. are the redundent cycles becoming
redundent cycles? Anyway, i'm guessing that something's not able to
hold the lid down on my bitbucket well enough and my bits are spilling
out all over the inside my cpu case. If anyone knows of a halfway
decent Firewire interface that can consistantly play nice with an 8500
without springing leaks everywhere, I'd love to know about it. 


thanx.