Is this possible?

[ymike673]

I was wondering if it's possible to adapt the Colecovision Steering wheel for the 5200. Would seem like a great way to play Pole Position.

[atariksi]

I was wondering if it's possible to adapt the Colecovision Steering wheel for the 5200. Would seem like a great way to play Pole Position.

 

If it's just a big potentiometer, you can use a multimeter and measure the resistance at the extremes and let us know and then it'll be easy to tell if it can be used as a big paddle or perhaps a small range of the steering wheel can be used. The Atari 2600 Paddle is like 0 kOhms ... 600+ kOhms.

[DanBoris]

I was wondering if it's possible to adapt the Colecovision Steering wheel for the 5200. Would seem like a great way to play Pole Position.

 

Not easily. The joystick inputs on the 5200 are analog, but the Colecovision steering controller is completely digital. Instead of a potentiometer it uses a digital encoder disk to sense the rotation.

 

Dan

[+Allan]

I was wondering if it's possible to adapt the Colecovision Steering wheel for the 5200. Would seem like a great way to play Pole Position.

 

Not easily. The joystick inputs on the 5200 are analog, but the Colecovision steering controller is completely digital. Instead of a potentiometer it uses a digital encoder disk to sense the rotation.

 

Dan

The trackball uses two digital encoder disks. One for left and right and one for up and down. If you can move the left and right encoder so that it lines up with a horizontal disk that is attached to the shaft of the steering wheel, it should work. Luckily Pole Position was written to work with both the joystick and the trackball.

 

Allan

[CV Gus]

It would be fairly difficult.

 

I built a paddle controller for the 5200. When you use it for the game Pole Position, it acts like a "real" steering wheel: turn it a bit, you turn a bit. Turn it some more, and you turn some more. This is why a digital controller is so awkward here: you either turn all the way left or right (hence, the skidding).

 

This, a paddle controller, works because the 5200 is able to understand signals from a potentiometer (variable resistor).

 

(To a 5200, there is no such thing as "neutral." The stick turns what are two tiny paddle controllers in the controller: one for vertical, one for horizontal. With, say, a 2600 joystick, if you let it go, then it is not sending any signals.

But with a 5200 controller, the joystick is ALWAYS sending a signal. "Neutral" simply means a resistance that the game programmers are telling the game not to react to; any less, go left and/or up. More=right and/or down. This is why, when you unplug the controller, it is as if you jammed the stick down and to the right; the circuits running from the controller port have been disconnected, causing an "infinite" resistance.)

 

A ColecoVision is NOT. It simply uses "pulse" signals; each time you turn a steering module or roller ball, it is as if you are "tapping" a joystick. You cannot really use a potentiometer for a CV, since it cannot understand anything but YES or NO.

 

If you position a paddle in Super Breakout in a certain place with either a paddle or 5200 joystick, turn the game off, and then start it up again, the paddle is in the same place, because the reading from the controller would be the same. This could not be done on such a game for the CV; the paddle controller I designed is more like the one from 2600 Indy 500- it just keeps turning. It would work, but never the same way as one for the 5200.

 

If you wanted to make such a controller, you would have to attach a sort of paddle controller to the shaft of the steering wheel, but- this is important- remember that a paddle controller cannot keep turning over and over; you'd have to do something to limit the way the steering wheel can turn.

 

An exception to this- if you wanted to do it this way- is to have the wheel of the potentiometer "leaning" (with a spring, or something) against the shaft. This way, even if you keep turning the wheel, you cannot break the potentiometer, even though it cannot itself turn any longer.

 

It can be done, but NOT with the existing hardware in the steering controller. You'd only use the "shell."

Edited December 2, 2008 by CV Gus

[+Allan]

It would be fairly difficult.

 

I built a paddle controller for the 5200. When you use it for the game Pole Position, it acts like a "real" steering wheel: turn it a bit, you turn a bit. Turn it some more, and you turn some more. This is why a digital controller is so awkward here: you either turn all the way left or right (hence, the skidding).

 

This, a paddle controller, works because the 5200 is able to understand signals from a potentiometer (variable resistor).

 

(To a 5200, there is no such thing as "neutral." The stick turns what are two tiny paddle controllers in the controller: one for vertical, one for horizontal. With, say, a 2600 joystick, if you let it go, then it is not sending any signals.

But with a 5200 controller, the joystick is ALWAYS sending a signal. "Neutral" simply means a resistance that the game programmers are telling the game not to react to; any less, go left and/or up. More=right and/or down. This is why, when you unplug the controller, it is as if you jammed the stick down and to the right; the circuits running from the controller port have been disconnected, causing an "infinite" resistance.)

 

A ColecoVision is NOT. It simply uses "pulse" signals; each time you turn a steering module or roller ball, it is as if you are "tapping" a joystick. You cannot really use a potentiometer for a CV, since it cannot understand anything but YES or NO.

 

If you position a paddle in Super Breakout in a certain place with either a paddle or 5200 joystick, turn the game off, and then start it up again, the paddle is in the same place, because the reading from the controller would be the same. This could not be done on such a game for the CV; the paddle controller I designed is more like the one from 2600 Indy 500- it just keeps turning. It would work, but never the same way as one for the 5200.

 

If you wanted to make such a controller, you would have to attach a sort of paddle controller to the shaft of the steering wheel, but- this is important- remember that a paddle controller cannot keep turning over and over; you'd have to do something to limit the way the steering wheel can turn.

 

An exception to this- if you wanted to do it this way- is to have the wheel of the potentiometer "leaning" (with a spring, or something) against the shaft. This way, even if you keep turning the wheel, you cannot break the potentiometer, even though it cannot itself turn any longer.

 

It can be done, but NOT with the existing hardware in the steering controller. You'd only use the "shell."

 

Pole Position has two controller modes. Joystick and trackball mode. The trackball works like the Colecovision trackball in that it uses 'Pulse' signals. If you use a 5200 trackball with the trackball cord, you can convert one to use a steering wheel. Open both of them up and you will see what I mean.

 

Allan

[DanBoris]

The trackball works like the Colecovision trackball in that it uses 'Pulse' signals.

Allan

 

Are you referring to the 5200 trackball? The 5200 trackball does use pulse encoding internally, but the output of the trackball to the controller port is strictly analog. The speed of the balls rotation is converted into a voltage level that goes to the 5200 stick input so moving the ball faster is equivallent to push the stick further in one direction.

 

Dan

[BigO]

The trackball works like the Colecovision trackball in that it uses 'Pulse' signals.

Allan

 

Are you referring to the 5200 trackball? The 5200 trackball does use pulse encoding internally, but the output of the trackball to the controller port is strictly analog. The speed of the balls rotation is converted into a voltage level that goes to the 5200 stick input so moving the ball faster is equivallent to push the stick further in one direction.

 

Dan

 

I haven't looked in detail at the CV steering wheel, but I think it outputs the pulses directly. I've done some analysis and repair on the 5200 trackball: there's a fair amount of circuitry in there to convert the pulse rate to a corresponding analog voltage (more likely current) as DanB described.

 

The way I see the circuitry in the trackball controller, it's a tachometer. The main difference is that with a tach, the output would be zero with no rotation and in the trackball the DC output is offset upward such that no rotation outputs, instead of zero, a voltage that is equivalent to "center" on the joystick pot(s).

 

So, even if you made an adapter (as I was originally thinking) to allow the steering module to mimic the 5200 trackball analog output, it would only steer left and right while the wheel was turning and would return to "center" as soon as you stopped moving the wheel.

 

I've done a bit of experimenting with reading gray code using a PIC uController. In theory, I think I could read the output of the steering controller (not sure of the "protocol" it uses) and vary an analog output based on the information conveyed by the pulses. The analog signal would stay where it is until more pulses are detected which would increase or decrease the signal according to which way the wheel was turned. This would create an output that would closely mimic a potentiometer or a "non-self-centering" version of the 5200 trackball output.

 

I'm working on a project that implements something similar so maybe I'll tear into one of my driving modules and see what it would take to build an adapter for the 5200 when I get that far. Not a big moneymaker, I'm sure.

[CV Gus]

The trackball works like the Colecovision trackball in that it uses 'Pulse' signals.

Allan

 

Are you referring to the 5200 trackball? The 5200 trackball does use pulse encoding internally, but the output of the trackball to the controller port is strictly analog. The speed of the balls rotation is converted into a voltage level that goes to the 5200 stick input so moving the ball faster is equivallent to push the stick further in one direction.

 

Dan

 

I haven't looked in detail at the CV steering wheel, but I think it outputs the pulses directly. I've done some analysis and repair on the 5200 trackball: there's a fair amount of circuitry in there to convert the pulse rate to a corresponding analog voltage (more likely current) as DanB described.

 

The way I see the circuitry in the trackball controller, it's a tachometer. The main difference is that with a tach, the output would be zero with no rotation and in the trackball the DC output is offset upward such that no rotation outputs, instead of zero, a voltage that is equivalent to "center" on the joystick pot(s).

 

So, even if you made an adapter (as I was originally thinking) to allow the steering module to mimic the 5200 trackball analog output, it would only steer left and right while the wheel was turning and would return to "center" as soon as you stopped moving the wheel.

 

I've done a bit of experimenting with reading gray code using a PIC uController. In theory, I think I could read the output of the steering controller (not sure of the "protocol" it uses) and vary an analog output based on the information conveyed by the pulses. The analog signal would stay where it is until more pulses are detected which would increase or decrease the signal according to which way the wheel was turned. This would create an output that would closely mimic a potentiometer or a "non-self-centering" version of the 5200 trackball output.

 

I'm working on a project that implements something similar so maybe I'll tear into one of my driving modules and see what it would take to build an adapter for the 5200 when I get that far. Not a big moneymaker, I'm sure.

 

 

I've built a prototype "no battery needed" steering mechanism. It's so crude it's laughable, but it does show that such a thing is possible. It's based on the design of the spinner of the SAC.

 

Unlike a CV, the 5200 is meant mainly to read analog. A CV? Digital. Therefore, they speak two entirely different languages; this is why my digital 5200 controller cannot work with a CV, and a CV controller cannot work on a 5200.

 

Look at it this way: 5200 and a hypothetical CV Pole Position.

 

The car in Pole Position, unlike the ones in Turbo and Pitstop, actually "turns." The ones in Turbo and Pitstop slide left and right, just as in those ancient electromechanical games.

 

So- if you wanted to program Pole Position for the CV, and use the steering module, how would you do it? Well, you could, and the car would turn just as in the 5200 version.

 

But the WAY it would be done is entirely different. With the 5200, it reads the resistance in the controller. Left=less=turn the car left. Right=more=turn the car right. But it's entirely based on the RESISTANCE of the controller: the more you move the controller, the more (or less) resistance there is, and the more the car turns. In other words, how much the car turns is directly related to how much (or little) resistance there is with the controller. If you knew the exact resistance for when the car first turns a bit to the left, and you pulled apart the controller and just soldered in a fixed resistor to those two wires, the car would be in that position. This is why, when starting a 5200 game, if you want the car to face straight ahead, the stick (or paddle) must be in the "middle position."

 

With a 5200, the HARDWARE- the controller- decides how the car is positioned. Just like the paddle in Super Breakout.

 

With a CV, though, it would be different. When you start the game, the SOFTWARE, not the hardware (as in the 5200) would make your car face straight ahead. After this, when you turn the wheel a bit to the left, the CV reads the order in which the "switches" are closed (this is why the holes in the CV Steering mechanism are uneven), and determines that you have turned the wheel a bit to the left. It would make the car turn a bit to the left. Turn it some more, or in the other direction, and it would act accordingly. It's absolutely identical in effect as if you had programmed it to turn the car one way or the other if you tap the joystick- tap it fast, it'll turn fast; tap it slow, it'll turn slow. Each tap turns it a bit.

 

The only way to directly use a CV Steering controller for a 5200 is if you add something to the hardware- a translator- that would, with each pulse from the controller, make it into a higher or lower resistance in the 5200 circuit. But remember that the 5200 controller is analog; the CV, digital. They could not understand one another.

Edited December 6, 2008 by CV Gus

[+Allan]

The trackball works like the Colecovision trackball in that it uses 'Pulse' signals.

Allan

 

Are you referring to the 5200 trackball? The 5200 trackball does use pulse encoding internally, but the output of the trackball to the controller port is strictly analog. The speed of the balls rotation is converted into a voltage level that goes to the 5200 stick input so moving the ball faster is equivallent to push the stick further in one direction.

 

Dan

 

You may be right. I didn't look at the programming end of it. It still should work as long as the 'disc' that was used to create the pulse signals was the same as the two that are used in the 5200 trackball. It's possible that the Colecovision trackball pulse disc won't work with the 5200 trackball pulse 'reader' because of the difference in size and position of the holes on the discs but I'm not sure since I haven't tried it.

 

Allan

[BigO]

Unlike a CV, the 5200 is meant mainly to read analog. A CV? Digital. Therefore, they speak two entirely different languages; this is why my digital 5200 controller cannot work with a CV, and a CV controller cannot work on a 5200.

[...]

The only way to directly use a CV Steering controller for a 5200 is if you add something to the hardware- a translator- that would, with each pulse from the controller, make it into a higher or lower resistance in the 5200 circuit. But remember that the 5200 controller is analog; the CV, digital. They could not understand one another.

 

Your view seems kinda narrow to me. It is true that the standard controller uses analog (potentiometers) to implement the man-machine interface for player positioning. However, the 5200 port has effectively 9 digital I/O and 2 analog inputs. I don't see why it couldn't read gray code or similarly encoded digital data as comes from a "spinner" type control. The 2600 has fewer digital I/O lines along side its 2 analog inputs at the controller port, but I imagine that most people consider it to be primarily a digital control method. There is no reason, other than the requirements of the game itself, that the 12 key keypad has to be part of a controller if you're designing one from scratch for a new game.

 

While I don't yet know exactly what sort of digitally encoded signal the CV driving controller outputs, I'd guess there's better than a 75% chance that a 5200 game could be written that would read the digital data stream. Assuming some reasonably intelligible sequence coming from the CV driving controller, the biggest trick would probably be to sample the port at a high enough rate to not miss any of the pulse sequence.

 

The "only way to directly use a CV steering controller for a 5200" that you describe is what I had just described when I said: "In theory, I think I could read the output of the steering controller (not sure of the "protocol" it uses) and vary an analog output based on the information conveyed by the pulses. The analog signal would stay where it is until more pulses are detected which would increase or decrease the signal according to which way the wheel was turned. This would create an output that would closely mimic a potentiometer or a "non-self-centering" version of the 5200 trackball output." I personally wouldn't consider such an adapter to be a "direct" means of using the CV steering controller.

 

Tangentially: The 5200 trackball controller connects to the same analog inputs as the standard controller. However, it does not actually provide a variable resistance. It outputs a voltage (as can be observed on an oscilloscope) that is proportional to the speed and direction that the ball is being spun.

 

I agree that the analog aspect of a 5200 controller couldn't be directly utilized by the INTV (or the CV), but it seems highly likely that the 5200 could work with a more digitally inclined controller if such a thing were desired.

[bohoki]

dont know what kinds of cars you all driv but my sterring wheel goes counter clockwise to turn the wheels left then it hits a stop,then i turn clockwise for right and it hits a stop

 

the only car i can think a person drives where the wheel goes around and around is a bumper car which it rotates 360

 

a pc wheel should be easy to find then you can use the capacitor trick to make the 5200 think the 100k is 500k

[BigO]

dont know what kinds of cars you all driv but my sterring wheel goes counter clockwise to turn the wheels left then it hits a stop,then i turn clockwise for right and it hits a stop

 

the only car i can think a person drives where the wheel goes around and around is a bumper car which it rotates 360

 

a pc wheel should be easy to find then you can use the capacitor trick to make the 5200 think the 100k is 500k

I think this conversation started with a discussion of the existing CV steering wheel expansion module which doesn't use a potentiometer.

 

To get a potentiometer based steering controller, the mods that people do to make a 5200 paddle controller from a 2600 paddle controller would be perfect. Just doesn't look like a steering wheel.

[CV Gus]

There are traditionally two kinds of trackball controllers: one that acts like a 5200 controller (potentiometer), and the "pulse" kind, like a CV. Wico made such. A steering module is just one with a wheel and only one plane of movement.

 

You know how the potentiometer kind works; paddles, just like in Super Breakout. The pulse kind acts like you're tapping a joystick really fast. The former is analog; the latter, digital.

 

If you wanted to build a pulse controller for the 5200, in theory it could be done, but indirectly. For example, the "switches" activated could be for controller buttons 4 and 6; each time the sequence is 4-6, the game would read that as a "left" move. 6-4 would be considered "right." The game programming would read this as with my "CV Pole Position."

 

If you check the spinner of a ColecoVision Super Action Controller, you'll see two reed switches that are closed by the two magnets on the wheel. If the sequence is left/left and right/right, then the move is, say, to the left.

 

If it was right/right and left/left, then the game goes the other way. The actual workings are the same as the Roller Controller or Steering module, but each spin does not work as many times, which is why using such a set-up for Turbo would be slower. The magnets can NOT be too close together. Now, look at the holes in the wheel inside the steering controller. Many more close/opens.

 

But to build a pulse controller for an EXISTING game, like Pole Position...tricky. It would have to be something that would make each pulse lower or raise the current going to the 5200, since that's how the game itself reads the instructions. But I don't think it could be done directly; the game reads an ANALOG signal; pulses are digital.

 

Your best bet is to use potentiometers. I would press a potentiometer against the shaft of the steering module, hard enough to turn, but loose enough so if you spin past the potentiometer's end, it won't break. A spring, perhaps. Maybe add a rubber collar to the steering column, for a better hold.

 

If you wanted to play it safe, you could add something (like a peg) to the shaft, so the wheel itself could not turn past certain points. But then there must be no slippage of the potentiometer. Cogs?

Edited December 16, 2008 by CV Gus

[+Allan]

Just so nobody is assuming at what's in the 5200 trackball controller I'm posting a picture so you can see the insides.

 

Allan

Edited December 16, 2008 by Allan

[DanBoris]

Just so nobody is assuming at what's in the 5200 trackball controller I'm posting a picture so you can see the insides.

 

Allan

 

You can also see another picture of the inside on my site along with the schematics:

 

http://www.atarihq.com/danb/52trakball.htm

[kenfused]

If you wanted to build a pulse controller for the 5200, in theory it could be done, but indirectly. For example, the "switches" activated could be for controller buttons 4 and 6; each time the sequence is 4-6, the game would read that as a "left" move. 6-4 would be considered "right." The game programming would read this as with my "CV Pole Position."

Does anyone know how fast most of the buttons on the 5200 can be read?

 

One fire button is a normal trigger input and can be read fast, but the other fire button and the keypad are all a keyboard scanned by POKEY and I don't know how fast the keyboard scan takes.

[Gorf]

A CX-22 could work on a CV as it too is all digital. All you would need to do a paddle

on a CV is and A to D converter...hook the A to the paddle and the D to the joyPort(s).

Of course you would need a few more things than this but it certainly could be done.

 

I dont remember how many signals per joy on CV but you can do probably a 6 bit

AtoD. Then you need games written for it.

[BigO]

But to build a pulse controller for an EXISTING game, like Pole Position...tricky. It would have to be something that would make each pulse lower or raise the current going to the 5200, since that's how the game itself reads the instructions. But I don't think it could be done directly; the game reads an ANALOG signal; pulses are digital.

I'm relatively certain that's what I said (twice) before, "[...] I could read the output of the steering controller [...] and vary an analog output based on the information conveyed by the pulses. The analog signal would stay where it is until more pulses are detected which would increase or decrease the signal according to which way the wheel was turned. This would create an output that would closely mimic a potentiometer or a "non-self-centering" version of the 5200 trackball output."

 

The adapter would perform the same sort of pulse to analog signal conversion that goes on inside the 5200 trackball with the key exception being that the analog output wouldn't "return to zero" in the absence of steering input, it would "hold" until additional pulses directed it higher or lower to produce the effect of the output tracking with the physical rotation of the steering wheel. I don't think it would be all that tough. I've already decoded the raw pulses (internal) from a WICO trackball with a dead chip and wrote PIC microcontroller code to make the trackball put out 2600 compatible digital joystick signals as it originally did. It needed some further tweaking, but once I get the proof of concept working, I tend to lose interest. That same or similar decoding logic could be employed and used to control a D/A converter of some sort to make the CV steering controller work with the 5200 as described above.

 

Similarly, one could "slow down" the pulse train and produce an output that would be compatible with the 2600 driving controller. I also proved out that concept with moderate success in my trackball-revival-uController project, but playing Indy 500 with a trackball is not an experience that I'd care to repeat.

 

When I get my logic analyzer that I ordered today, I should try to analyze the I/O of a CV steering controller (provided mine works...I haven't ever used it).

Edited December 19, 2008 by BigO

[midnight8]

I was wondering if it's possible to adapt the Colecovision Steering wheel for the 5200. Would seem like a great way to play Pole Position.

 

Old analog pc steering wheels are a very easy mod to the 5200. I made one myself. Its actually a flight simulator controller. The buttons are the gas and brake It does work great on pole position. A unique way to play super breakout too.

Edited December 31, 2008 by midnight8