Making Star Wars Trilogy force feedback hardware work

[Bart]

Oh no That's really too bad. I'm not too familiar with SWT's force feedback? Did it vibrate the joystick or actually knock it around in different directions?

[luisfcorreia]

Good news!

the board has been found!

SWT-ffb.jpg (244.89 KiB) Viewed 7861 times

SWT has force feedback in two axes, that's the big difference from all other force feedback machines.

ffb-hw.jpg (140.72 KiB) Viewed 7861 times

Even more news as I'll uncover them!

Cheers!

[luisfcorreia]

aaaand after a HUGE facepalm, I see that this driveboard ROM is present in the game files since forever.

it's what epr21119.ic8 is for, drive board rom

now on with investigating supermodel3 innards

[SailorSat]

First tests on my side didn't show any feedback outputs on the usual ports - lamps worked fine.
I'll check in MAME later.

[luisfcorreia]

AKAICS, in MAME the doesn't even work properly.

I tried MAME 179 and I only get a black screen.

[Joaquim Gonçalves]

That Rust...



Can´t wait to see some miracle in the future.

[SailorSat]

AKAICS, in MAME the doesn't even work properly.
I tried MAME 179 and I only get a black screen.

I can boot the game, but I am unable to reach the service menu for now

There seems to be some hardware access to 0xC10000xx

1 write to 0xC1000038, 0xC0 and two reads from 0xC1000014 (checked for 0x01 and 0x04) -

What model3 stepping does sw trilogy use?

[luisfcorreia]

that rust will be dealt with.

we've got a plan be to make FFB work again even if through some other techniques.

BTW, did anyone bother with figuring out how the drive board talks to the actual power board?

[Bart]

AKAICS, in MAME the doesn't even work properly.
I tried MAME 179 and I only get a black screen.

I can boot the game, but I am unable to reach the service menu for now

There seems to be some hardware access to 0xC10000xx

1 write to 0xC1000038, 0xC0 and two reads from 0xC1000014 (checked for 0x01 and 0x04) -

What model3 stepping does sw trilogy use?

This is the SCSI controller range but as far as I know, the Step 2.x games (like SWT) do *not* have this device. The Real3D firmware libraries were almost certainly written to be compatible with multiple board revisions, hence why they often seem to contain code that can use either the SCSI for DMA transfer or the custom DMA device that appeared in Step 2.x.

This sequence of writes is unusual. It is not checking device ID but rather looks like it's setting the device up and then checking whether any interrupts are pending. Is this what happens in Supermodel? I don't think I take any precautions to make the SCSI invisible on Step 2.x games. But Star Wars definitely uses the custom DMA device (which we emulate in Real3D.cpp) for all transfers between CPU<->Real3D memory.

DMA should have nothing to do with the drive board.

[SailorSat]

Hm...
We'll find the feedback data - should not be too complicated.

I am curious because the usual drive board port does not seem to be accessed at all. Lights are accessed as usual.
I doubt they did something special - an 8 bit port should be more than enough