Rotary controller, raphnet usb adapter, Tempest 2000 and mame

[joeyaage]

Has anyone tried a rotary controller with the raphnet usb adapter? My windows 8 PC will not run vj. I think it may be because of amd processor or graphics. However, vj does work on Debian linux on same PC ?!?!

I have not tested raphnet adapter on linux yet but think it may work.

 

Also, Jaguar rotary controller would be excellent for mame tempest/arkanoid/etc. The problem is there is no mouse mode a la stelladapter. Would it be possible to write a driver for mame to enable usage of the rotary controller?

 

Any thoughts?

[Trooper Galactus]

Actually, has anybody thought of trying to port a version of Arkanoid to the Jaguar? Be nice to get another rotary controller game, and that's one of my faves! Have the arcade version in a cocktail cabinet and the NES version CIB!

[Zerosquare]

Actually, has anybody thought of trying to port a version of Arkanoid to the Jaguar?

You mean like this or this? Edited May 14, 2017 by Zerosquare

[Trooper Galactus]

You mean like this or this?

If only I could actually find a copy of Impulse X...but has anybody made an Arkanoid cart?

[+CyranoJ]

If only I could actually find a copy of Impulse X...but has anybody made an Arkanoid cart?

 

Yes, I have

[Trooper Galactus]

 

Yes, I have

Would like.

[+CyranoJ]

Would like.

 

Yes, you probably would

[joeyaage]

What with the increase in rotary controller games and the new Jaguar USB tap on Kickstarter, it got me thinking about this topic again. I have carried out some tests on a Windows 7 laptop with the Raphnet adapter. When using the game controller properties from Windows control panel it can be seen that for each small turn of the rotary ( a 'step') the data produced is - right turn: +x,-x, 0, +x, -x, 0, +x, etc. and left turn: -x, +x, 0, -x, +x, 0, -x, etc. From what I remember reading about rotary controllers this is the correct behaviour and to be expected. Good news. :-)

However, when trying the controller with virtual jaguar it DOES NOT work for playing. :-(

I have not looked at the source code for virtual jaguar yet. Perhaps it could be modified for rotary input. I tried Arkanoid 2 (the old version with no sound) the bat would move to the right (not very smoothly) but not to the left. Left turn would just jitter left/right. Kobyashi Maru, the spaceship would just jitter and could not be controlled. It is a bit like the effect when you try and use a rotary controller as a standard dpad without reading and interpreting rotary data. I tried full circuit rocketeer and managed to get the jet pack to move a bit to the right but left would not budge just jittering. The great Tempest 2000 would just jitter, however I tried it with project tempest and had a bit more success. The claw would move left and right quite well but fire (the B button) would now not work ?!!?!. Curious.

 

 

Has anyone tried using a rotary with the new Jaguar USB tap? There is PCB schematics and firmware/driver source code on GitHub. Perhaps the source code could be modified so that the rotary acted as a mouse device instead of a game controller device. If that was achieved then the rotary could be used together with the Atari Vault program (on steam and to be included on the new Atari VCS). There is already code in place to play Super Breakout, Pong, (Major Havoc and Tempest I am not sure about) and the 2600 paddle games with a mouse.

 

Also perhaps the source could be changed so that we have a rotary acting as a game controller that would work with ANY left/right game. Imagine this with classic space invaders or Jeff Minter's new Tempest 4000.

[Gummy Bear]

You might just have to...

 

 

 

Buy a Jag!

[+CyranoJ]

Shamus can jump down my throat and tell me I'm wrong but if I had to guess it would be...

 

VJ probably only reads the controller once a frame and passes it to the state machine. Rotaries need to be scanned at a >1frame resolution.

[joeyaage]

Ah yes, that would contribute to the errors. The first thing I thought of was that maybe the input was read (for a left turn): -x,0,-x,0, etc. And for right a right turn: +x,0,+x,0,etc. This being on a Jagpad. So that when it reads the rotary it goes (for a left): -x,0,+x,0,0,0,-x,0,+x,0,0,etc. Or something like that. In other words it is processing the release state as well. I am trying to visualise what effect that would have but it would cause an error. It would kind of depend on how the input code was constructed.

Hopefully I will get a chance to look at the source for vj. It might makes things slightly clearer.

[Zerosquare]

There are two things that may be issues:

 

- as CJ said, rotaries have to be read much more frequently than standard controllers to work fine. That being said, it should still work if you turn the rotary really slowly.

 

- the pattern for rotaries goes like this:

(clockwise / counter-clockwise may be swapped, I'd have to check)

 

Note the "both" state, where both the left and right directions are active at the same time. This is only possible if your adapter exposes the D-pad as individual buttons, not as X and Y axes. In the latter case, rotaries won't work properly.

Edited June 4, 2018 by Zerosquare

[Machine]

If only I could actually find a copy of Impulse X...but has anybody made an Arkanoid cart?

 

 

 

Yes, I have

 

Any plans to release this? I would love another rotary game!

[jamm]

On 6/4/2018 at 9:34 AM, Zerosquare said:

 

Note the "both" state, where both the left and right directions are active at the same time. This is only possible if your adapter exposes the D-pad as individual buttons, not as X and Y axes. In the latter case, rotaries won't work properly.

 

I'm surprised there hasn't been more talk on this subject.  I just tried the Raphnet adapter in Windows and looked at the built-in gamepad configuration/calibration utility.  As I turn the dial clockwise, I see the axis cursor move from the right-most position, to the middle, to the left-most position, and then back to the right-most position again.

 

I'll attach a GIF of the action to this message.

 

Anyway, it seems that the adapter is presenting the controller as an axis, and not as individual buttons.

 

Edited February 18, 2020 by jamm

[Zerosquare]

You should ask Raphnet about it. It's something he can easily change in his code.

[jamm]

14 hours ago, Zerosquare said:

You should ask Raphnet about it. It's something he can easily change in his code.

I did ask him about it over email about a week ago.  Haven't heard anything back, but I'll update here if I do.

[Spider-Dan]

I've e-mailed Raphael a couple of times about controller-related questions (even once with a $10 PayPal donation attached!) and he's never responded.  So...

[jamm]

He did respond once to an email I sent him about it.  He said he could change the firmware, but didn't have a rotary controller to test with.  I offered to mail him mine, but I never got a response to that.