Right Turn Clyde - wheel problem

[slider2732]

 Simple controller problem, my steering wheel only goes right in MAME.
There are two wires (green and blue) that exit the Sega built controller section and that usually run to the wiring harness of Top Speed. Top Speed can make sense of it, MAME seemingly can't.

Am quite used to mouse hacks and even did a Happs trackball to PC conversion last year, but, either way I connect the 2 output wires, it goes right. Tried 2 different mice and actually the Happs hack itself by using the left/right wires..same deal, right only.
However, there do seem to be varying amounts of movement...and I do have a few spare mice, so will try a few more in case it's a 'gaps to light pulses' issue.

I would guess that Sega didn't build the original controller.
Anyone got any tips ?

[Angry_Radish]

Is this an optical or analog wheel?
Either way, don't you need 3 wires instead of 2?

[slider2732]

It's an optical, sorry not to mention.

The pinouts for Chase HQ (which I believe Top Speed follows) show a Centre connection on the harness for the wheel....but the wiring to the controller on the wheel is +5V, Ground, Green, Blue.

Interestingly, connecting the mouse feed affects the 2 LED's on the controller. Putting diodes on the outputs for 'output only' stops the feedback and restores the LED's to lighting for direction. However, I can't simply hack to the LED's and do it that way, cos they stay lit without wheel movement. Sometimes they are both lit, but it's not predictable, sometimes they flash around too.
When running the real arcade PCB the LED's do the same thing.

[slider2732]

Oh, and, my mouse hacks in the past have only needed 2 wires (to the outer 2 of the IR reciever for left/right direction.
I'd have thought 3 as well, but it seems that's what the mouse already sorts out from the pulses.

[slider2732]

FIXED! 

Hopefully of use to others who may have the same troubles -
The key to this was in the sensing. Normal mice have 1 IR receive for left/right and 1 for up/down. It would seem the logic onboard detects what does what for each receiver and that's how come the spacings in the holes of the plastic wheels get to be important.
So....went down to my basement and found an older mouse. It has seperate sensors on it for each direction. EX and RX were marked on the circuit board and tallied to senders and receivers, a total of 8 LED looking things.
I found which were the common 2 connections for the left/right and that left the data pulse 2 wires.
Soldered connections to these and YES! it works.
It was the wrong way around at first but that's sods law lol
The steering wheel is actually more responsive than I thought it would be...a driving pleasure

 

[slider2732]

 
The steering now doesn't work at all in Top Speed.

Perfect in anything else i've tried but not that one and THAT is the cabinets game!!!
Guess I might be looking at emulation for other games and running Top Speed's PCB for real LOL

[u_rebelscum]

My guess: the wheel outputs direction and speed (mice and optipac expect a quadratic signal).

Normal mice have 1 IR receive for left/right and 1 for up/down.

Err, normal mice have both sensors in the same unit; they save pennies by having one emitter (LED) and two sensors in one unit.

It would seem the logic onboard detects what does what for each receiver and that's how come the spacings in the holes of the plastic wheels get to be important.

The spacing is important so the sensors take turns being blocked and not; problems happen if both sensors change at the same time.

The steering now doesn't work at all in Top Speed.

You mean in mame, right?  What about the real PCB?

Mame is treating the input as a POT wheel (270 degree).  Did the game have two types of steering wheels?   Not uncommon, but should be noted if it does.

The wheel should be able to work in mame anyway; just map the wheel to the AD Stick analog input, and NOT the AD Stick inc/dec, nor the joy left/right/up/down (which is a hack to let keyboards work).  I'll test tonight.

[u_rebelscum]

Okay, it took longer than a night but ...
Somthing is weird is going on with the driver & mame core.  Digital devices (keyboard, 8-way sticks) work great (well, as well as a digital input can do at faking an analog input).  Absolute analog devices (POT steering wheel, analog joystick, ect) usually work fine turning right, and sometimes work turning left. Relative analog (mice, spinners, trackballs) can only turn right.  Other games that originally had absolute analog inputs, such as SpyHunter, play okay with all three types. 

I have a couple guesses what's going on, but with the core input changes Coming Soon, probably not worth working on it.  My top guess ATM: the minimum is set to 0xff7f & the maximum to 0x80.  The decimal, these should correspond to -129 minimum and 128 maximum (in signed 16 bit), but 0xff7f could be seen as 65407 in unsigned 16 bit, or any 17+ bit.  I think mame is thinking the min/max is -65407/128 since 65407 > 128. 

A short time hack is to edit the topspeed.c driver and set the min,max to 0x81,0x80;  it seems to work on my computer.

Or you can go back and use an old version of mame (0.112u2 works, 0.113 doesn't).

[slider2732]

Thanks
Yep, I know what you mean about the mice...the 4 leg output receivers just seemingly less common to myself. Older HP (with Logitech innards) and those *ahem* that are usually found in Minnesota Clean Up Week are my 'usual' Must have about 2 dozen lol

The Top Speed now steers. By looking at the DIP's in Mame I found that I hadn't selected Analogue Steering in a Deluxe Sitdown Cab. Um, yeah, well mine is upright...but that setting worked !

Gears are still screwy. I soldered the spare leg on the shifter microswitch, so that on/off was reversed and now games seem to be the correct way around. Interestingly enough Top Speed included..which I would presume, with verification, could be a call for a dev to mod the driver file.
I'm not a dev in any way and hats off to those that are.