Best mouse to use for a steering wheel hack

[leapinlew]

I have an Atari Steering wheel from a 1976 Atari Lemans game. I've tried using a mouse hack on the wheel, but when I install it - I have to spin the wheel several times to get the mouse from one side of the screen to the other. I've cranked up the mouse sensitivity in Windows and it then it moves faster, but kind of skips around.

I have a couple thoughts on this - please let me know if I'm thinking in the right direction.

Does the windows sensitivy even matter? Are the setting in mame more effective at controlling the sensitivity.

Is there a brand of mouse better to hack than another brand? Do some mice have better software or resolution?

Any help will be appreciated. Thanks!

[Kremmit]

How many teeth are on the encoder wheel?  I've seen some older arcade parts with really, really low-res encoder wheels.  If you only pass 10 teeth per revolution of the wheel, and the mouse is expecting to see 40 teeth pass it's optics for every inch you move the mouse, then you've got trouble.  (I just pulled those numbers out of my butt, but you get the point.)

Also, which type of mouse hack are you doing:  Are you passing the encoder wheel through the mouse's optics, or are you wiring the optic board from the LeMans wheel to the mouse PCB?  If you're trying to pass the LeMans encoder wheel (big) through the mouse's optics (really tiny), they may be incompatible.  In order for an encoder wheel to work with a set of optics, the spacing between teeth has to correlate with the spacing between the optics just right, or the optics won't properly register every tooth, or even get the direction the wheel is spinning right.

In theory, the Windows mouse settings don't have anything to do with MAME.  In reality, sometimes they seem to anyway.  Generally, it's preferred to disable "Enhanced Pointer Precision", "Mouse / Cursor Accelleration", and drop the mouse sensetivity all the way to the bottom, so as to remove Windows from the equation, as much as is possible.

Some mice do indeed have better resolution.  I've got this really neat old Logitech serial mouse that I love, but when I plug it in to a Windows 95 machine (it's incompatible with anything higher; it's that old), even with the Windows sensetivity jacked all the way up, you've got to drag it all the way across the mouse pad a few times to get across the screen.  My old Razer Boomslang was so sensitive at it's highest setting that even a tiny hand tremor would send the pointer halfway across the screen.

The mechinical parts can provide higher or lower resolution depending on the number of teeth on the mouse's encoder wheels, and the number of revolutions of the encoder wheels for every inch the mouse is moved.  That last bit is determined by the size of the mouse ball and the rollers it turns, which in turn spin the encoder wheel.  Higher quality optics can miss fewer encoder teeth, and keep up with a faster spinning wheel.  A better controller chip can keep up with faster pulses from the optics.

As for software, software can't truly increase resolution, but software can apply it's own "Mouse Accelleration", before Windows gets the signal.  It can increase the polling rate on the PS/2 port.  Better written software might do a better job of keeping up with the signal from a fast-spinnng encoder wheel.  And probably some other stuff I don't know about.  High-fallutin' mouse drivers, like the ones that come with gaming mice, might do some mouse accelleration, which you don't want.  I'd just stick with Windows generic drivers when possible.

Some of that actually matters, and some of it isn't likely to make much difference.  I'd look at your encoder wheel and the type of hack you're doing first, before you go shopping for a better mouse.

[SirPeale]

Also, don't forget that when you're in GAME things will be different.  Yes, the mouse is expecting to see those 40 teeth per inch, and will report to Windows as such.  Games will be different.  When I was testing trackballs I would note that spinning the trackball in a circle would give me a *tiny* circle.  Games pretty much worked okay, though still needed a bit of tweaking for optimum performance.

[leapinlew]

Kremmit + Peale -

Thanks a ton for the information. This is good info and I realize I didn't provide anywhere near enough information.

As for the mouse hack... silly me, I thought there was only one. I took a mouse apart and I am running the Lemans teeth(really big) between the mouse optics (which were really small). I haven't secured it so I could try it in a game - I just noticed it took a lot to move it across screen. I'm going to count the teeth when I get home.

I'll stick with the generic drivers.

Last question: If this doesn't provide enough turning capabilities in the game, as in I have to spin the wheel 3 full revolutions to turn in the game, what are my options? Does someone sell a board? Could I print black lines on a transperency sheet of paper to give me the proper resolution?

[u_rebelscum]

... If you're trying to pass the LeMans encoder wheel (big) through the mouse's optics (really tiny), they may be incompatible.  In order for an encoder wheel to work with a set of optics, the spacing between teeth has to correlate with the spacing between the optics just right, or the optics won't properly register every tooth, or even get the direction the wheel is spinning right.

For best results, if tooth width = gap width = X, and distance between sensors = Y:

Y = X * (n + 0.5)  if Y > X where n is a whole integer >= 1, or
X > Y  (as long as the teeth fit in between the light source & sensor)

Changing the X or Y so that n switches from odd to even (or to X>Y) switches the direction the mouse thinks it's spinning.  (Odd to odd, even to even or even to X>Y don't change direction.)

The worst case is Y = X * n, where n is a whole integer >= 1.  The sensors can't tell which direction the wheel is turning.  The closer to this, it's easier for the sensors or mouse to make a mistake.

Of course, sometimes tooth width <> gap width, so that complicates above equations.

Anyways, in your situtation, X > Y and it works, so you should be good AFA the sensors working correctly.  Having enough resolution per rotation is another story.

In theory, the Windows mouse settings don't have anything to do with MAME.  In reality, sometimes they seem to anyway.  Generally, it's preferred to disable "Enhanced Pointer Precision", "Mouse / Cursor Accelleration", and drop the mouse sensetivity all the way to the bottom, so as to remove Windows from the equation, as much as is possible.

Also, don't forget that when you're in GAME things will be different.  Yes, the mouse is expecting to see those 40 teeth per inch, and will report to Windows as such.  Games will be different.  When I was testing trackballs I would note that spinning the trackball in a circle would give me a *tiny* circle.  Games pretty much worked okay, though still needed a bit of tweaking for optimum performance.

These two are partially related, in that the windows settings shouldn't effect mame.

Beyond that, the original game might be expecting a specific teeth/rotation ratio than what works well in windows.  Mame has an "analog sensitivity" (ingame, tab, analog controls, xxx sensitivity) that can be changed on a per game setting.   The defaults usually are for translating "standard" windows mouse resolutions to what the game expects (IOW, < 100%).  Since you don't have a standard window resolution mouse with your hack, you can increase this setting to 100% or up to 255%.  But as Kemmit mentioned, any setting > 100% "increases" the resolution in software: a 20 teeth @ 200% is choppier than 40 teeth @ 100%.  Increasing the mame sensitivity to 100% won't hurt you, and will help a little but maybe not enough.


If you want, you can try making a higher res (more teeth/gaps) encoder wheel with transparencies.  Others here have made it work, and if your search you might find some encoder wheels to download.  I'm not sure if they match you wheel, though.

[Kremmit]

GroovyGameGear's "Trackball Booster Kit" consists of a pair of fairly high-res encoder wheels.  You'd probably have to drill the center hole out to attach one, but that might be an option for a replacement encoder wheel.

But there's no guarantee those would work with the mouse's optics, either.  Doing the other mouse hack might work better than messing with the wheel, unless it is really, really low-res.

[horseboy]

GroovyGameGear's "Trackball Booster Kit" consists of a pair of fairly high-res encoder wheels.  You'd probably have to drill the center hole out to attach one, but that might be an option for a replacement encoder wheel.

They come with a hole.

[Kremmit]

GroovyGameGear's "Trackball Booster Kit" consists of a pair of fairly high-res encoder wheels.  You'd probably have to drill the center hole out to attach one, but that might be an option for a replacement encoder wheel.

They come with a hole.

Yeah, but it's really small.  I suppose it *might* be the right size for the screw on his LeMans wheel.. but then again it might not be.

[leapinlew]

Thanks for the info guys -

I'll let you know how it turns out. I didn't take too many pics of this project, but I'll make an announcement anyways.