CD32 controller pad issues

[rjd324]

I realised that when I booted up Payback the "blue" button to get into vehicles was not working. So, I tried to test out the controller using a program called "Controller Test" (https://aminet.net/package/util/misc/ControllerTest).


I have tested this controller on various systems (EDIT: By test I mean that I have tested using "Controller Test"):



- A600 vampire 2 (coffin r57): ensured that cd32 mode is selected for the joystick port: directions are fine, but all buttons are mapped to different buttons and pressing blue also triggers another button simultaneously. Digital mode is okay.


-A500 ACA500+ACA1233n (3.9 bb4): everything works in this machine. (*EDIT*: Actually, this is not true; see below)


I then tested it on 3 different A1200s (2*2b 1*1d4, 3.9 bb1 or 2):
-Everything works except BLUE never registers. (*EDIT*: The controllers do work on the 1200 motherboard with or without an accelerator card)


The controller is one of those third party CD32 controllers.


I do not even care too much since I do not really play games. The only thing I want to know is why this is happening; what matters to me is ensuring that the problem is not with my amiga hardware.


I have not changed lowlevel.library and they all seem to be using the same default one since the sizes are the same.



Can anyone offer advice?

[jotd]

You can try a better tester: http://aminet.net/package/driver/input/JoyPortTest.lha

It can use lowlevel.library or not, which is good because on some motherboards lowlevel doesn't work that great.

you can also try whdload games that support extra buttons. The button routine is pretty robust to all controllers.

[rjd324]

So the issue I have with JoyPortTest is that it is not detecting all the buttons when pressed using the CD32 controller. This happens on both the A500 (which you can recall from my original post is the one that reads all the CD32 buttons correctly using ControllerTest) and the A600.

Interestingly, I tested Desert Strike on both machines. Latest history version from the ReadMe file:

 History:
 --------
 
 12.may.2019 - V3.1 by Keith Krellwitz/Abaddon & JOTD:

I select nothing in the initial pop-up WHDLoad menu.

On the A600, when I press the Blue/X button the menu comes up (as if you pressed F10). Then, immediately exiting the menu the weapon fires. As if that button is wired to two lines.

On the A500, when I press the Red/A button the same happens as above.

If someone can test this version of that game out using a CD32 controller and let me know what happens then perhaps we can say that it is the controller that is causing the issue. Afterall, it was just a cheap replica CD32. Plus, I find it hard to believe that all A1200 motherboards would exhibit the same fault (again, see my initial post).

Can someone test it?

Regards.

[jotd]

Your pad is faulty. The Desert Strike joypad read code is the best there is. Better than lowlevel library, fixed by hordes of people. Plus there's 2 button joystick / joypad detect at startup

Your joypad seems to pass as a real joypad, but then doesn't work properly / has timing issues when reading the buttons.

[rjd324]

Okay, despite what I said before about the Blue button not registering on the A1200s I actually tested DStrike on one of my 2Bs and it was fine.

To be clear, this is the controller I am using https://lyonsden.net/wordpress_s/wp-...4-1024x768.jpg (This is just a screenshot that I found. I bought mine from ebay and I know that inside it is using a PIC chip, not the LS shift register).

I found that:
Left shoulder -> nothing
Right shoulder -> cycle weapons
Blue (X) -> cycle weapons
Green (Y) -> nothing
Yellow (B) -> menu
Red (A) -> fire weapon
Start -> menu

If you can confirm that you get the same results then that means this controller works on my A1200 (I did not test the other two 1200s) but behaves strangely on my A500 and A600 even though the A500's ContollerTest actually seems to indicate each button press correctly.

What do you think this means?

[jotd]

readme of Desert Strike:

	Red 	- Always fires the Cannon
	Blue	- Always fires Hydra
	Green	- Always Fires Hellfire

 	Red 			- Fire
	Yellow 			- Bring Up Map Screen/Exit Map screen
 	Play 			- Pause 
 	Play+Forward+Reverse 	- Quit
 	Green+Forward+Reverse 	- Abort Game

so what you describe looks normal to me. You could try green+forward+reverse just to see if it quits to wb to test left shoulder (reverse) and green.

As said earlier, whdload slaves with joypad controls use the "state of the art" joypad reading routines. try others, I've fixed dozens of them last year. I don't think I can help more.

[rjd324]

I have some interesting results for the A500 machine.
When I use this (https://ibb.co/37nFKsp) configuration DStrike works fine

But, when I use this (https://ibb.co/khKyqJc the difference simply being that the 1233n is enabled compared to the first picture) I get the issues I mentioned earlier; One button being mapped to both the Menu and Fire at the same time.

it seems that the accelerator is somehow affecting things and presumably the vampire is affecting things on the A600.

I know that you cannot help more, jotd, but in case anyone else is interested... at least, these are my findings.

[rjd324]

I wonder if it could be a combination of the way the PIC chip is programmed and the clockin signal being faster when using the 68030.
https://ibb.co/p2bKkYD

[MrD]

(I received your email ) Yeah that seems possible to me, depending on how the PIC chip is programmed, if the faster clock speed is making some transitions get missed. A full photo of both sides of the pad pcb would be useful otherwise we're just guessing.

[rjd324]

As for the A1200, I tested it both with and without an accelerator card.

When there was no accelerator card, I was getting about 35-40KHz input to pin 6/_FIRE1 (pin 11 CIA) and with the accelerator card there was no difference in frequency and DStrike worked on both configurations. On both the A600 and A500 with a vampire and 1233n respectively, those input frequencies pretty much doubled and the controller malfunctions.

I guess it would be interesting to find out why _FIRE1 has different output frequencies in the 500/600 with those accelerators. Perhaps that warrants a different thread.

I am waiting for my shift register to arrive, and I might try and modify the "CD32" controller I have here using the shift register to see if that makes a difference. I also wonder if I can bypass the pin 6 input using an arduino to provide some programmed PWM at about that freq of 35-40KHz...

[Aardvark]

Tried this? http://aminet.net/package/game/patch/SetJoyPort

[rjd324]

Thanks for the tip, but I still get the same issues.

I am going to wait for some of the SN* chips to arrive to build a CD32 controller using the "official" shift register technique too see what happens.

[rjd324]

I received a prototype CD32 controller board (https://www.ebay.co.uk/itm/KMTech-Am...72.m2749.l2649) and applied all of the components. Across all my machines I tested it out and it was consistent. Though, for some reason I have to hold R1 if I want GREEN/YELLOW to register (basically any button other than RED and BLUE and UP/DOWN/LEFT/RIGHT). I will check soldering later. But, the real reason I bought that thing was just to see if it was the PIC chip causing the issue. The fact that it is consistent is kind of what I was looking for.

Things that I have noticed is that all of the controller tester software never seems to work properly for CD 32 controllers, at least it does not for me. Not criticising the software; thankful for it!

I am now quite skeptical about getting any controllers that do not use the shift register to implement CD32. There are a number of apparent "honeybees" out there that are just fakes using PIC controllers.

Also, does this https://amigastore.eu/414-rys-mkii-usb-adapter.html allow me to use a compatible controller for it that will allow CD32 buttons? Let's say a SNES controller, will I have green/red/yellow/blue?
Regards,

[DanyPPC]

I own the RysMk2 and I use it with a standard snes usb controller (INNEXT).
RysMk2 support CD32 emulation as well as 1 button and 2 buttons joystick.

It works even on A1200 rev. 2b, the more problematic revision with CD32 controllers.

[rjd324]

Okay, so I ended up getting a CD32 gamepad controller and this thing just works fine.

The RHYS and all the other controllers using the PIC chips (or w/e) cause issues.

Guess I am going to have to look more into this, but at least I have a controller that works now.