rec.autos.simulators

Even after disabling "Poll with interrupts" the problem still
persisted. I then discovered the PDPL4 gameport card that is digital
and is not affected my these analog frequencies that affect the
Standard Gameport.

But then I ran into another problem , my motherboard Abit KA7100
wasn't compatiable with my GeForce2 card,, the machine would not POST
with this card but worked great with a V3, the GeForce worked great in
the older Celeron 400a system,  so I couldn't run N4/GPL in all its
glory..but I digress.

So I recently built a SocketA based AMD 900Mhz system, with the
downturn in prices I couldn't resist (can you say 60 bucks for a
900Mhz CPU?)  I can  !!

Using a  Asus board this time (A7V-E) and I sent the Abit board RMA to
be repaired

Used the same case and power supply (HIC 300GR) and used a PCI based
SB 16 soundcard. My joystick port glitch problem seemed to lesson  but
in GPL and NASCAR4 when racing I"d noticed that after a few laps the
precission my my wheel slowly went away.. after about 5 laps steering
became almost impossible..especially on straights, if I went back and
recalibrated it was finally for another 5 laps or so..

I could no longer use the PDPL4 because the new mobo (like most new
mobos, didn't support ISA) I got a RockFire USB joystick port adapter
this finally resolved the glitches in the wheel but whats odd is when
I calibrate the wheel in N4 or GPL the max values for steering are
only 0 to 90, where it should be 0 to 255. But whats more interesting
is I don't noticed any difference in my steering while racing...
Hmmmmmm,, actually turning faster laps now..

Perhaps the values are incorrect and are actually more then just 90,
but my question is why hasn't this problem been addressed by a PCI
based joystick card. Are the big companies not aware of this problem?
Am I the only person with this problem?

Also the only piece of hardware that is the same from the Abit based
system and the Asus based system is the power supply, perhaps it is
the culprit in all of this.....

When I calibrate in N4/GPL with the rockfire device the calibration
needle is rock solid when using the gameport on the soundcard the
needle  jumps around a bit...

And I recently install new POTs in my TSW so I think they are fine..

I have looked into that problem a bit and it seems like the Rockfire adapter
is designed for pots with higher values, around 250 kohm. I don't know the
exact maximum pot value, since I haven't tested it.

At the max pot value you will get the highest possible readout (255) in
games and also the highest possible resolution.  The Rockfire adapter has
only about 100 separaste positions to report within the 0-255 range. This
gives that at that if you use a 100k pot, there is only about 30-40
positions reported from the adapter to the games.

There is two possible solutions for this problem. You can replace your pots
to bigger ones, or modify the Rockfire adapter to report higher values at
lower potentiometer resistance to give you the full range. The later method
involves replacing a capasitor for each axis. I haven't documented the
operation, but I can look into my adapter and tell more you about the
capasitors to replace.

The problem with the Rockire adapter are due to the calibration method used
for it. It is the operative system/games that gets calibrated to the adapter
and not the adapter to the potentiometers.

I guess not, and it is strange that it isn't more talked about.

--

Mauritz Lindqvist Lindqvist
Sweden

Spamguard: Remove .ajabaja in e-mail

If you absolutely, positively cannot find it, get back to me; I *might* be
able to unearth the info...but then I'd have to kill you.

--Steve Smith

http://www.thomas-superwheel.com/prods/usbadp.html

snip:

Improves stability of calibration, at the cost of slightly lower axis
resolution

On Fri, 21 Dec 2001 19:27:05 GMT, "Stephen Smith"

I guess you used the ECCI wheel with the PDPI card? There is some registry
edits you can do to optimize the range, but those are specific for the PDPI
card and has nothing to do with Direct Input itself. There is some general
type of regedits for DInput, but I think they can't increase the reported
axis resolution for the Rockfire USB adapter.

Take a look at the graphs I have put together to see the differance between
the regedit method and modifying the hardware. The graphs show both
calibrated data and raw data. Raw data is reported directly from the device
without any modifications from DInput. The red dots represents every
reported position from the axis.

http://w1.111.telia.com/~u11103812/Rockfire.jpg

The tested axis has a resistor range of 0-35 kOhm.

But I would really like to know if there there is any regedit to do, to
change the resolution for the USB adapter.

--

M Lindqvist

1. USB is slower than an IRS refund.

2. If you can find one, the PDPI is still the best gameport interface ever
devised by man.  (Not that that was the original issue.)

--Steve

Yes, the Reg edits were for the PDPI, but I was under the impression that
there were Reg values that accomodated the range of any analog I/O device
(that would be converted from 0-40 or 0-250,000 ohms to the traditional
digital values of 0-256).  I could be terribly wrong.

Guess I have no right to take a life, huh?

--Steve