Only 2 Buttons and D-Pad Work on CD32 Joypad

[fondpondforever]

Hello, i just bought an official amiga cd32 joypad from eBay and only 2 buttons and the D-Pad are working. The Red, Blue, Up, Down, Left and Right ones. This seems absurd as the Green, Yellow and other buttons do not respond. Loaded up a CD32 Gamer Demo a few hours ago and wanted to play Tower Assualt by pressing the Green Button and it didn't work. Pressed the Red button which loaded up Super Stardust. Don't know if there's something wrong with the Controller or Console. Blew the sockets a few times because i thought it might be dusty but that didn't do anything. Any help would be much appreciated. Thank you.

[xArtx]

There’s a logic chip the CD32 controller that shifts out the extra buttons... it doesn’t sound absurd.
It’s got to be either the chip itself, the clock signal from the machine to the controller,
or the data back to the machine from the controller (or anywhere in the plugs/sockets/cable in that path).
Power and ground connections would also have to be seen by the chip in the controller.

[roy bates]

if you can send me the joypad i can fix it for you.
just send me a pm for my address and add your for return postage...simple.
its fairly easy to fix.ill let you know what it was after ive fixed it...

its up to you,you wont get too many offers lke this

[cpiac64]

insert an audio cd for test all buttons

[Daedalus]

Nice offer there from Roy I'd also offer to repair it but he got in there first! Yes, it's not that complicated a controller but it does depend on power and the clock signal from the machine, whereas a basic joystick requires neither. So one of those lines or the chip in the controller are acting up.

Have you got another controller to test the socket with?

[fondpondforever]

Thanks for your help, I'll send it to you Roy to fix if that's ok. Haven't got another Controller to test it with at the moment but I'll try the audio cd idea you mentioned cpiac64.

[roy bates]

no worries,ive sent my address via pm

[fondpondforever]

Thanks.

[wiz12]

Sorry to revive this old thread but I have an original CD32 pad showing similar behaviour to the OP.

Using JoyPortTest - green, yellow, start and the two shoulder pads don't respond however if I untick lowlevel.library 47.1 then all the buttons will work. Is this normal behaviour for original CD32 pads or is there a possible problem with it?

[Daedalus]

It shouldn't be normal behaviour, no. Lowlevel.library does some detection to automatically determine the correct controller type so perhaps this isn't working. It's hard-wired in the controller though so I don't see how that part wouldn't be working and the other buttons are working fine (as you can see by bypassing the library) unless there's a bad connection on some of the pins of the chip (or the chip itself is faulty).

Also, you definitely have lowlevel.library installed? It's not present as standard on 3.0 and below.

[wiz12]

I'm using OS 3.2.2.1 and have version 47.1 of lowlevel.library in the libs directory. Maybe there's an alternative version of this library that wouldn't have this issue?

I have since tested a second original CD32 pad and this behaves similarly to the first pad in that green, yellow, start and the two shoulder pads don't respond however if I hold down a button, right for example, and then press any of those non-responding buttons, they then respond.

The main difference with this pad however is that if lowlevel.library is unticked, those buttons do not respond at all, which is the opposite of the first pad.

I guess it's possible the chips on these could be faulty. Is it possible to find replacement chips and would they need to be programmed?

[Daedalus]

There's always the possibility that the behaviour has changed in version 47.1. The version that is standard on the CD32 and with OS 3.1 is 40.x. It might be worth swapping that version in and checking if anything changes.

If the chips are faulty, they're readily available. The standard CD32 pad uses off-the-shelf logic chips so no problem at all to replace them.

It's very odd that the D-pad changes the behaviour of the buttons though since they're totally independent of the D-pad. It makes me wonder if there's some other issue like a broken trace, damaged component or poor connection on the pad or even the motherboard.

[wiz12]

I tried version 40.35 of lowlevel.library which I think was the one included with WB3.1 but it didn't change the behaviour. I also tested both pads on another motherboard but both pads behave in the same manner which leads me to believe there's some fault with the pads themselves.

Though it is only one of the CD32 pads where holding down a direction button whilst pressing either green, yellow, start or the two shoulder pads makes those buttons respond when they normally won't which is quite strange.

[Retro1234]

Have you tried a game?

On my A1200 the CD32 pads were so sensitive if they weren't 110% pushed in tight they went into 2 button mode.

[wiz12]

I tried Dangerous Streets and the pad where you have to hold down a direction to activate one of the top coloured buttons is definitely faulty, the other pad seems to work fine.

2 button mode on a CD32 pad? I didn't know this was possible?

[Daedalus]

Pushing in the connector only part way isn't an A1200-specific issue. If pin 5 isn't connected, it will revert to 2-button mode and this happens on any Amiga model. The A1200's ports aren't any different from standard in the many A1200s I've seen and I've never had any issues with bad connections and CD32 pads on any A1200s, though as always Commodore have used different parts depending on what was cheapest at the time. I have seen A500s that were missing pins though that caused issues with some controllers and not with others. This was down to the owner having a joystick with a connector in which a pin had previously broken off. Using that stick on other ports then pushed that pin in too far, making it difficult to get a good connection with any other controller.

In this case it does indeed sound like a controller issue. If pin 5 (or the associated wire in the cable) is damaged, it will work as a 2-button pad for backwards compatibility. The lowlevel.library detection routine will also fail to recognise it as a CD32 pad in this case.

[wiz12]

I tested for continuity on each wire from port end to gamepad PCB end and the blue wire or pin 6 was disconnected from the PCB. However even after re-soldering this it again made no difference though I have noticed something that might be unusual in JoyPortTest.

With lowlevel.library selected when pressing any of the non-responsive buttons like green or yellow "Ctrl type" shows "n/a" whereas the working buttons show "joypad".

With lowlevel.library not selected, the working buttons show "joypad" whereas the non-working buttons change from "joypad" to "joystick" when pressed. The strange thing is the on-screen blue boxes do not detect a key-press or highlight.

I guess maybe one of the chips could be faulty to cause this?

[Daedalus]

Pin 6 disconnected would also have the effect of preventing it from working, though it would also stop the red button from working in 2-button mode. Yep, could be a chip issue, could be a fracture in a PCB trace or pad. The information from JoyPortTest might be enough to determine exactly what's causing the issue, but without knowing internally how the program works I can't determine any more.

[wiz12]

I'm thinking it could be a faulty chip as I can't see any broken traces or obvious signs of damage. I'll have to experiment more with it. Thanks for your help Daedalus.