16 replies to this topic

[Legend]

I was wondering if there was any kind of in depth faq about the atari 2600 paddle controllers? I'm particular interested in a certain point.

I grew up with the 2600 when I was small. But It was eventually discarded and I don't have exact memories of how the paddles worked. A few years ago, I got a set of paddles to use with my flashback 2 and stelladaptor. When I got them, it seems that the paddle is only responsive for about a quarter of it's full range of motion. It still moves the paddle on screen all the way across and back, but if I turn it to the left, it keeps turning after the on screen paddle stops and I have to keep turning back to the right until it reaches the point that the onscreen paddle stopped before it starts to move the paddle again. Which makes it so that if I turn too far to the left, I"m basically going off screen and have to catch up to get back on screen. This can be quite frustrating when playing something frantic like Kaboom. Is this the way they are supposed to be?

I expected that when I turn the paddle all the way to the right, it would be on the right side of the screen. And if I turn it all the way to the left, the on screen paddle would reach the left side at the same time that the paddle in my hand is turned completely to the left.

I also have a Jakks tv games paddle set that works the way I just described. So the actual paddles and the tv games one both handle very differently.

I hope that all made sense.

Edited by Legend, Thu Jul 15, 2010 11:20 PM.

[Mitch]

The way your paddles are working is correct. It really depends on how the game was programmed. Going from memory it seems like most games did not use all of the paddles range.

Mitch

[Legend]

Mitch, on Thu Jul 15, 2010 11:28 PM, said:

The way your paddles are working is correct. It really depends on how the game was programmed. Going from memory it seems like most games did not use all of the paddles range.

Mitch

It seems that all the games I've tried don't use the full range. It's like there is a 75% "deadzone on the left side. Do you know any games that for sure use the whole range so I could test it?

[Rybags]

I think by nature the paddles will have a deadzone, but it should be something like 10% or less on one that's working well.

No idea if it's the Pot itself returning a constant resistance, or the computer reading that range of voltages as one value (probably the more likely case)

[tetrode kink]

As others have stated, your paddles are working correctly.

I think the reason it works this way is because back in 1977 it would have been too expensive to include analog-to-digital converters in the VCS circuit design.

Nowadays, AtoD conversion is a dime-a-dozen, and it would read the paddle position in the background. The hardware would just assign a number from 0-255 to the current paddle knob position. All the programmer would have to do is read the number whenever he wants to see where the knob is, then position the character accordingly.

On the VCS it's a completely different story. The programmer actually has to write code to charge a capacitor, wait a bit, then read the register that represents the capacitor's state to see if it has discharged yet. This waiting-and-reading sequence has to be repeated dozens or hundreds of times in game code just to get one read of the paddle's position. The time it takes to discharge the capacitor determines where the paddle knob is positioned. The thing is, it takes an achingly long time to discharge the cap, about two full video frames' worth. This amounts to 1/30 of a second, which doesn't sound like much, but in game code 1/30-second might as well be forever. If the programmer can afford the code-execution time to completely discharge the cap, the game will make use of the whole paddle.

Unfortunately, reality makes this almost never possible. Considering the program spends about 3/4 of its time just keeping the screen alive, it's not surprising that most (all?) paddle games only devote enough cap-watching time to make use of part of the paddle's range. Also, perhaps because the VCS's paddle-handling hardware is so primitive, programmers came up with different code schemes for reading the paddle. This is why different games use different areas/ranges of the paddle for game input.

Interestingly, there has been talk here of workarounds to this limitation. This thread, started by tremoloman2006, asked if there was a way to make a paddle less sensitive so that it used more (or all) of its range. That spurred quite a response, in which several ideas were discussed. So far there is no solution which would use a paddle's range for all games. But there were several ideas (including schematics) for a down-and-dirty solution involving modding the paddle with a few parts so that it doubled the paddle range for any game, with a switch included so that the paddle could be switched to "original" mode when desired.

HTH,
-tet

Edited by tetrode kink, Fri Jul 16, 2010 2:09 AM.

[A.J. Franzman]

I tested a bunch of paddle games some years ago and didn't find any that use more than half of the control's physical range. However, in many paddle games, speed is more important than accuracy, so making the controller use the full ~300 degrees of rotation will actually worsen performance (or at least tire out the player faster) in those games.

Edited by A.J. Franzman, Fri Jul 16, 2010 2:10 AM.

[Legend]

A.J. Franzman, on Fri Jul 16, 2010 2:09 AM, said:

I tested a bunch of paddle games some years ago and didn't find any that use more than half of the control's physical range. However, in many paddle games, speed is more important than accuracy, so making the controller use the full ~300 degrees of rotation will actually worsen performance (or at least tire out the player faster) in those games.

Quote

I think by nature the paddles will have a deadzone, but it should be something like 10% or less on one that's working well.

That's interesting. So it seems they all have dead zones but, of a different amount. So perhaps if I picked up another set sometime, it might have a bit more physical range but still not the entire range?

Did each paddle have a different amount of responsive range A.J.?

[A.J. Franzman]

I did not test a large number of paddle controllers, I only used one set. I also only used one console. I don't know what the parts tolerances are for the pots in the controllers, the capacitors on the VCS motherboard, or the TIA's read/discharge circuit, but I'm sure that if you add them all up it's got to be fairly large. I believe there is also a design difference in the consoles; some of them have fixed resistors in series with the paddle lines and others don't.

[Nukey Shay]

Just a side note:
The "dead zone" present for any given game may also be done through software. Breakout and Super Breakout both read the paddles for most of the scanlines...which would give a very wide range. However, the programs trim the value before using it. Super Breakout cuts away quite a bit, like a quarter or fifth of the range read IIRC, because 12 playfield pixel's worth of space is dead (6 on either side). Breakout only trims 2 pixels worth of space for the left side.

[Video]

It always bugged me that the paddles didn't use their full range of motion, but i imagine it has something to do with price of the post used, vs pots that would have been more proper. The pot in the 2600 reeds up to 1500 ohms (I think that's what you call it) while the atari only reads about 300 of those. You can actually replace the pot with a smaller ohms version, and it'll work on a fuller range, but being used to the old pots, I never could get used to it.

The one good thing about the limited range of the pot read, is that it was very easy to build an analog stick out of one (though the only game it works with is Marble Craze)

[Legend]

I wonder if something about it can be done in an emulator like stella to adjust the amount of "deadzone" On the paddles when using stelladaptor?

[Nukey Shay]

But then it would be less authentic compared to original hardware. The dead areas are due to the game program itself. You could hack a paddle controller to replace the potentiometer with one that has less resistance...that would provide more turning radius without the dead spots appearing so quick.

[JingleJoe]

Just thought I'd pipe up because I am doing/have done alot of testing in this area
No Atari 2600 paddle compatible game I have tested so far uses the whole range and most only about ¹/₃ of the entire range. Breakout uses a slightly larger range than Super Breakout and it varies slightly between games.

The potentiometers used in the paddles are 1M (1,000,000 ohms) and the range value in ohms for most games was about 30k to 350k I think, however different games use different ranges, for example I think Breakout used about 50k to 400k, whereas many others seemed to get almost the full range with just a 220k pot suggesting thier range is somewhere in the region of 30k to 250k. It varies, but when using a 1M pot, games never seems to use more than about ¹/₃ or ²/₅ of the whole range.

[Rex Dart]

Legend, on Thu Jul 15, 2010 11:18 PM, said:

Which makes it so that if I turn too far to the left, I"m basically going off screen and have to catch up to get back on screen. This can be quite frustrating when playing something frantic like Kaboom. Is this the way they are supposed to be?

I expected that when I turn the paddle all the way to the right, it would be on the right side of the screen. And if I turn it all the way to the left, the on screen paddle would reach the left side at the same time that the paddle in my hand is turned completely to the left.

Yeah, the thing with turning them too far is that you're turning them too far. Don't do it

I imagine using the full turning range of those controllers would cause crazy wrist strain before too long. I'd prefer the feeling of adjusting a tuner dial to the feeling of turning on a faucet any day.

[Nukey Shay]

No doubt...but you could help that issue by switching the paddle caps with ones from an old Pong console. It's a perfect fit, and the smaller dial is there if you need it.

[31336haxx0r]

As this is supposed to be the paddle FAQ, I have an entirely different question. The paddle schematic shows, that the paddle potentiometers aren't wired as voltage dividers but variable resistors instead. I suppose I'll only get correct paddle readings, if I use the correct 1 MOhm pot?

[zylon]

Having the paddles move to the stops would be a bad idea as hitting them frantically would damage them over time. I use C64 paddles or upgrade to 500k ohm pots like C64 used. Smaller range of hand motion required for the equivalent distance on screen.