Amiga CD32 joypad frustrations

Discussion in 'Commodore - C64 to Amiga CD32' started by phoenixdownita, Sep 1, 2013.

  1. phoenixdownita

    phoenixdownita Spirited Member

    Joined:
    Apr 29, 2012
    Messages:
    160
    Likes Received:
    8
    I recently bought a CD32 and I'm having a few issues with the joypad.

    The pad I have works great on the audio CD sections [all buttons and all directions] and in most games [Flick, Banshee, Alien Breed TC] but some just baffle me. In specific the Project X SE + UBB game disk does not seem to work. I can chose the game and start it by pressing the red button. But neither of said titles would recognize anything I press past that. On the same CD there's Super Frog demo which works just fine.
    The same problem with Alien Breed 3D, I cannot enter the security code because it does not recognize presses to the red button.
    [it is not the port, if I put an SMS pad I get dual button support and the games do recognize the presses so I can get the to start and play]

    I suspect that some games switch the joypad in Joystick mode and my pad doesn't like it much, I will do further research on that but I wonder if anyone else had any experience with them?

    I have the schematics http://gerdkautzmann.de/cd32gamepad/cd32gamepad.html and validated my pad matches [with the exception of the inductance on the output] ... also my pad came modded by having pin 2 of the LS165 cut and connected to P6, also pin 11 of the LS125 was cut and connected thru a diode to P6. Finally there was a resistor from J4 [+5V] to pin1 of LS165 and another resistor from GND to pin 11 of LS125 [soldered to the diode now on the pin].

    Anyway I modded it back to original spec [removed diode, resistors and reconnected pin2 LS165 and pin11 LS125 to the board] and it works on joypad mode [all buttons] but likely not in stick mode, then I put the mod back [diode and all] and have the same results, works in pad mode [all buttons] not in stick mode.
    When I say stick mode I mean Project X and UBB and AB3D don't register the red button presses.
    I validated all connections pin by pin from the cable to the connector and those are not to blame also all traces match up.

    I do have another joypad arriving in the next 2 or 3 weeks and hopefully it works fine so I can copy its circuit and fix this one, again most of the games seem to work fine but in case someone has a hunch/had to deal with it, please let me know.

    Any help greatly appreciated.
     
    Last edited: Dec 31, 2014
  2. ASSEMbler

    ASSEMbler Administrator Staff Member

    Joined:
    Mar 13, 2004
    Messages:
    19,394
    Likes Received:
    1,029
    Problematic you don't have another controller to test.

    Are you 100% sure the controller is not worn out or defective?
     
  3. phoenixdownita

    phoenixdownita Spirited Member

    Joined:
    Apr 29, 2012
    Messages:
    160
    Likes Received:
    8
    Most games work fine but some don't. There' really isn't much that can be worn out but it is possible that some of the capacitors/resistors are out of spec causing some timing issues.

    Case in point if I put UBB I need to connect a MD pad and press B to start the game, if I switch controller with the CD32 pad then it works until the end of the match but then again I cannot press Red to start another play.

    It is baffling.

    Obviously if I put an audio CD and go into the CD32 audio player I can press all the buttons to my heart content and it all works all the time.
    Instead those CD100/200/350.... work only on the menu but games that should work with a normal 1 button do not sense any press at all, if I connect the MD pad it all goes back to working.

    So for the most part is just plain annoying as this pad is supposed to be back compat on top of that for those 3 or 4 cd32 games [ProjectSE/UBB, Guardian, AB3D ...] it is just not working.

    I do not have available all of the replacement components [mainly the caps] but once I do I will replace them all just for staying on the safe side. I am not sure I care to attempt replacing both ICs [74LS165 and 74LS125], nothing too expensive just not sure the problem is there to start with.

    Oh well the controller was sold as is and it had been obviously mucked around, I guess the previous owner also got issues with it and couldn't get it fixed, it is also a challenge because the components involved are kind of simple in a way.

    EDIT1:
    I may actually have fixed my issues.

    I added a 1K resistor between VCC [P7] and PotX [P5] and now it works like a charm.
    What the mod does is to pull up [to VCC] P5 in case it is not been actively driven [neither GND nor VCC].
    I suspect that the way the CD32 works is by switching the pad back and forth between the 2 modes [joystick and joypad], it has to do so because only when it is in stick mode it can capture the buttons (the 74LS165 is set in LD), when it is in pad mode instead it shifts what it read while in stick mode and pipe it thru P9 one button at a time using P6 as the clock (the 74LS165 is set in SHIFT mode).
    In stick mode P6 is an output [the red button], in pad mode is an input [the clock].

    So I believe that in my case as the CD32 was switching from stick to pad and viceversa the timing got a little off during the Hi-Z mode, but by putting the pull up resistor there [between P5 and P7] I force the pad in stick mode during said time and that should buy enough time for the buttons to be read correctly ... at least this is my speculation.

    The point is that 1 single resistor of 1K soldered between pins 1 and 2 of the 74LS125 seems to be solving my issues as those 2 pins are connected to P5 and P7 on the cable.
    I tried Project X SE + UBB and both games seem to work [UBB recognizes the 4 buttons], also the CD100/200... now work both in the menu and in game [just tested Supercars 2 but it is encouraging]. Finally AB3D lets me enter the password and then I can play [it seems you fire with the green button in that game].

    Obviously I tested the CD32 audio player and all buttons work as expected.
    I will do some more testing to confirm that I'm out of the woods, I have to test some more of the CD32 games that were working just fine before the "fix" and some more 2 buttons games to make sure it is all as expected.

    I'm sharing this in case someone has similar/same issues with a CD32 pad.

    I am not sure if the issue can be caused by some other factor [maybe my CD32 port 1 is weak, maybe some of the pad components are out of spec and mess up the timing], in any case it is a relatively safe mod to do with no ill effect on the CD32. Not sure if a bigger resistor (10K) could work with the added benefit of even less power drained.
    The 1K resistor over 5V drains 5mA, with a power consumption of 25mW, I think it should be just fine, according to the datasheets the 74LS125 (http://susta.cz/fel/74/pdf/sn_74125.pdf) is ~15mA while the 74LS165 (http://susta.cz/fel/74/pdf/sn_74165.pdf) is ~30mA so the added 5mA is around only 10%.

    EDIT2:
    I've done some more testing and it seems it is working way better than before.

    All the CD32 titles that were giving me issues now work (Final Gate, Project X SE, UBB, Guardian, AB3D) and the others seem to be working as they were before (AB:TC, Flick ...).

    I still have some issue with few titles that I tried from the CD100/200... until I plug a real 2 button joypad they don't work quite right but that could be WHDLoad or some more pad timing issues as some other titles work.

    Anyway at least now I can play the CD32 versions of the games and that is what I wanted to start with, for those CD100/200... I have no issue if I have to plug in a real 2 buttons pad.

    EDIT3:
    A few more tests later and still happy camper.
    None of the CD32 proper CD based games seems to be having issues .... yeahhhh!!!

    Regarding the CDnnn compilations it seems to be more random and I am not sure it is worth investigating further.
    The CD100/CD200 seems to be fine on the handful of games I tried, the CD350/CD444 gave me some trouble instead.

    I will wait for the other controller to arrive and see how it fares and report for the rest of the community to see.
     
    Last edited: Dec 31, 2014
  4. dutchconsolefreak

    dutchconsolefreak Peppy Member

    Joined:
    Sep 8, 2005
    Messages:
    319
    Likes Received:
    4
    It has been a while i had anything to do with the Amiga/CD32 on the hardware side, but what i remember is that the CD32 didn't actually have a pad mode but instead a stick and paddle mode. The paddle mode worked by measuring the time it takes to discharge a capacitor influenced by the resistance of the potentiometer in the paddle.

    When they decided to use serial communication for the pad on the CD32, they kind of abused this cap-measuring line for signaling purposes during serial communication. I think it is imaginable that this capacitor is degrading over time, especially because the polling frequency is probably higher during serial comms than normally during paddle position measurement.

    I vaguely remember this cap was located inside the Amiga/CD32, but i'm not 100% sure.
     
  5. phoenixdownita

    phoenixdownita Spirited Member

    Joined:
    Apr 29, 2012
    Messages:
    160
    Likes Received:
    8
    Any chance you can find out which ones?
    I'd gladly replace them on my board [I already inverted the 2 backward caps that plague most PAL CD32s].

    Although that would not explain how the audio CD player works flawlessly detecting all buttons the way they are supposed to be .... anyway if you can find which caps on the motherboard could be the culprit it will greatly help.

    BTW you are correct in saying that the joypad mode is really a serial mode based on paddle mode .... I just gave it a nickname.
    Also having the same pin [P6] have to switch between in/out without any protection on the pad side is borderline criminal, I am pretty sure there' a current limiting resistor CD32 side [is there?] or it would be too prone to shorts.
    Commodore did a really poor job there, at least Sega used the Select line one way only.
     
    Last edited: Dec 31, 2014
  6. phoenixdownita

    phoenixdownita Spirited Member

    Joined:
    Apr 29, 2012
    Messages:
    160
    Likes Received:
    8
    Just received my second pad, after the obligatory cut of the plug and replacement with a new DB9 it all works like a charm.
    No issues with audio CD player, AB3D, Project X SE + UBB (both work fine) and I also tried Xenon 2 from CD200 with success.

    The PCB circuit seems exactly the same with the exception of a 6 holes ferrite bead series on P9, and 1 different value capacitor (221 rather than 220 and it's right there on the SH/LD line of the LS165, maybe this is the culprit as 1 order of magnitude can change the timing quite a bit).
    I'm still waiting for my resupply of capacitors to arrive then I will try a complete recap of the pad (it's just 5 of them). I also ordered the same 6 holes ferrite bead so I would be able to add that too, hopefully the 2 controllers would behave identically after the fix and end my frustration.

    Now if it all pans out and http://www.a1k.org/forum/showthread.php?t=32822 (German site, use Google translate, it does enough to allow you to understand) turns out to be cheap enough I may be enjoying my CD32 even more.
    Seriously, AB3D is too slow, I've been told that just by having fast RAM its performances are like 3 times faster.

    EDIT1:
    Got my capacitors today. Changed the 220 with a 221 and it seems all is fine and dandy, I've left the 1K resistor between pins 1 and 2 of the 74LS125 for now.
    Once I get the ferrite bead I will complete the mod, remove the extra resistor and report back. But for now it's all still good and the pad is alive and kicking.

    EDIT2:
    The supplier of ferrite beads messed up and I haven't yet received them so there's no updates on fixing my pad to be exactly like the working one. Hopefully I will receive them maybe in another 3 weeks!

    EDIT3:
    DEFEAT. :(

    I got my beads and I wired one up as per the working pad, so now the 2 pads looks identical (as far as I can tell at least).

    As soon as I removed the extra 1K resistor between pins 1 and 2 of the 74LS125, the pad went back to its old habits and doesn't work on AB3D password screen.
    So I put the extra resistor back and now it works again so I'll leave it in with the bead and put this whole issue to rest.

    It does bug the hell out of me though that one pad works fine and the other only works with the extra 1K R, it makes no sense at all but then again I do not have an oscilloscope to see the wave-forms to tell what is going on. I'll have to live with the doubt.

    I hope someone else would find this info useful for his own pad.
     
    Last edited: Dec 31, 2014
  7. RetroSwim

    RetroSwim <B>Site Supporter 2013</B><BR><B>Site Supporter 20

    Joined:
    Dec 10, 2012
    Messages:
    605
    Likes Received:
    26
    Based on your first post, it really seems like something fishy is going on with your system.

    It just happens that tonight I have knocked together the controller schematic as per this guy on a breadboard like so: http://i.imgur.com/MvnpQUu.jpg

    There are a few deviations:
    - I don't have an inductor on the output
    - My electronics store didn't have the schottky version of the shift register, so I used a 74HC165.
    - All the capacitor values I used were either rough guesses, or compromises based on what I have in my bits box. Looking at the schematic, I used:
    -- 22pF ceramics for C1 and C4
    -- 33uF electro at C2
    -- 4.7uF filter caps across the two ICs
    - I didn't have a resistor array for RA1, so I just used discrete 10k 1/2W resistors for pullups.
    - R5 is 220R, but it's just pulling down SER, I don't think the value is important.
    - Lastly, I added some LEDs to monitor the enable and clock lines.

    So far, I've tested Trolls and Pinball Fantasies, and the circuit works as expected. From watching the LEDs, I saw that Pinball Fantasies has a much higher clock rate than Trolls.

    What I've built is far from exact, but it is functioning properly on two games with different polling rates. I reckon your system might be playing silly buggers, rather than your controllers.

    Putting a resistor between pins 1 and 2 of the buffer IC creates a funny state.

    Normally, in Dpad mode, the A gate of the buffer should be in a low impedance state, this:
    - Prevents the pressing of the red button from interfering with the clock signal, by having gate D in a high impedance state.
    - Allowing pin 1 of the shift register to be driven high through RA1 by having gate B in a high impedance state.

    If you keep the A gate in a high impedance state:
    - Pins 4 and 13 of the buffer will be pulled low through R1, allowing their outputs to be driven high or low.
    - Pin 1 of the shift register is pulled low through the 0 state B gate.
    - The CLK signal from P6 will be affected (attenuated I guess) when the red button is pushed, since pin 11 of the buffer will be driven low whenever the button is pushed.

    This all being the case, I have no idea how your controller is working the way it is. Got an oscilloscope? ;)
     
  8. bart_simpson

    bart_simpson Dauntless Member

    Joined:
    May 13, 2011
    Messages:
    768
    Likes Received:
    18
    Do you have a official pad if so i fix one a while ago.
     
  9. phoenixdownita

    phoenixdownita Spirited Member

    Joined:
    Apr 29, 2012
    Messages:
    160
    Likes Received:
    8
    I used the same schematics and noticed that the wiring of my 2 controllers is now identical and all components have the same nominal values [slightly different than those reported but nonetheless the same across my 2 pads].
    I am not sure it is my system as the second pad works without a hitch, it always did, no modding or any other funny business..

    My pulling up A gate of the LS125 guarantees that pin3 is NOT driven as such it is pulled to GND by R1 allowing the RED button to be pass thru, but this is true IFF joystick pin 5 [P5] is left floating, if pin 5 is driven either high or low the 1K resistor does nothing at that point as it is the CD32 controller port doing the active driving of A gate of LS125.

    Anyway the other controller works just fine and all the components and traces are exactly the same, BTW the issue with the funny controller was never in DPad mode, it worked just fine in that mode, the issue arose on those games that switch it to normal stick mode momentarily, give AB3D a try and let me know. I had issues only in the original code check page then it worked flawlessly.

    [BTW I do not have an oscilloscope to see what the heck is going on]
     
    Last edited: Oct 23, 2013
  10. phoenixdownita

    phoenixdownita Spirited Member

    Joined:
    Apr 29, 2012
    Messages:
    160
    Likes Received:
    8
    Yes, I do have a second pad that never had issues. I am not sure why the 2 behave differently, it may be that my 74LS125 VCC filtering capacitor is a little out of spec and causing slight instability, not sure, I thought about replacing every single capacitor and resistor on the board just to see if it would fix it but never had the time to get around it and I am not too displeased with the 1K resistor
     
    Last edited: Oct 23, 2013
  11. RetroSwim

    RetroSwim <B>Site Supporter 2013</B><BR><B>Site Supporter 20

    Joined:
    Dec 10, 2012
    Messages:
    605
    Likes Received:
    26
    I initially thought so, but thought for a sec that maybe 1K was a low enough value to keep the voltage above the buffer's logic low, since all the pullups the design calls for are 10K. *shrugs* If it works, it works I guess.

    OK, I'll give Alien Breed a go.

    I notice that the clock LED goes dark during certain parts of Trolls' intro, not sure if that's significant.

    I'll do some more mucking around and see if there are any incompatibilities.
     
  12. RetroSwim

    RetroSwim <B>Site Supporter 2013</B><BR><B>Site Supporter 20

    Joined:
    Dec 10, 2012
    Messages:
    605
    Likes Received:
    26
    I can't get my CD32 to read CDRs, I've spent the last 2 hours adjusting the 4 trimmers on the PCB and the one on the laser, to no avail. :(

    I'll have to either find a real copy of Alien Breed, or a new laser for my system. Bummer.
     
  13. phoenixdownita

    phoenixdownita Spirited Member

    Joined:
    Apr 29, 2012
    Messages:
    160
    Likes Received:
    8
    Take your time. Another test albeit less reliable (as in 1 out of 10 it works anyway), is Project X SE + UBB disk.
    Both my controllers can navigate the selection menu and play Frogger demo, but my "flaky" one had issues with Project X SE, in specific I couldn't bring up the selection menu and start a game, once I pass that (by switching controller with a megadrive pad) I could plug "flaky" back in and play the game using both buttons (fire and weapon select) to my heart' content. No issue of course from the working pad or after putting the 1K R.

    I'm really interested in your findings although I do suspect some of my components in the "flaky" one may be close to limit tolerance hence the problem I am seeing.
    If we were to think about it logically the 1K R that I put in only has an effect during the intervals in which joy P5 is not driven, that means some of the Rs or Cs or the ICs is reacting too fast/too slow wrt perfectly in spec components, but without an oscilloscope I can't tell and yet the 1K R "fix" is much cheaper than anything I can think of (purchase of an oscilloscope included)
     
    Last edited: Dec 31, 2014
  14. RetroSwim

    RetroSwim <B>Site Supporter 2013</B><BR><B>Site Supporter 20

    Joined:
    Dec 10, 2012
    Messages:
    605
    Likes Received:
    26
    About a day after that post I ordered a new laser, and today it has arrived. So tonight I'll install it and hopefully get Alien Breed running.

    Fingers crossed! :)
     
  15. RetroSwim

    RetroSwim <B>Site Supporter 2013</B><BR><B>Site Supporter 20

    Joined:
    Dec 10, 2012
    Messages:
    605
    Likes Received:
    26
    Well I put the new laser in, and it's better than before, the Team 17 logo now comes up 100% of the time, nearly instantly. But it hangs there forever, with the CD access light on.

    I've tried a few different CDR brands, and two different burners. Any ideas?

    I'm going to try a different game, just for sanity's sake, but I'd really like to try AB3D.
     
  16. The Perfect K

    The Perfect K Robust Member

    Joined:
    May 11, 2011
    Messages:
    265
    Likes Received:
    42
    Regarding your controller problems - I assume this is the stock, super-wide CD32 controller? The one that looks all funky and is seen on the box? When I got into CD32 collecting, I ran across a post on some forum (possibly amibay? maybe the english amiga boards? I can't remember now off the top of my head) where someone claimed that the stock controller has trouble functioning as a 2 button joypad for legacy games. The gist of the post was that you should only use CD32 pads for CD32 games, as they'll exhibit the problems you are claiming. The post directed me further to the CD32 Competition Pro pads because they'll function as a 2-button joypad no problem.

    I have since never seen a single other mention of this problem until this topic. I actually don't have any stock CD32 pads to try it out (I do have UBB and Project X-SE) but I do have 2 CD32 Competition Pro pads and they work just fine under those games as a 2-button gamepad.


    Might I recommend picking up a amiga gravis gamepad? It's the best 2-button gamepad I've found on the Amiga, mainly because it maps both up and down to buttons 3 and 4.
     
  17. phoenixdownita

    phoenixdownita Spirited Member

    Joined:
    Apr 29, 2012
    Messages:
    160
    Likes Received:
    8
    Yes, here we are talking about the standard CD32 pad, the boomerang.

    After my experimenting with the one troubled controller the 1K Resistor on 74LS125 seems to be "fixing" just fine, the other CD32 pad never had issues and it works in 2 and 6 buttons mode/games without a hitch (at least that I can tell).
    I repeat, what is baffling me is that circuit wise the 2 controllers that I have are identical, the only thing different could be out of spec components, I do not have an oscilloscope to figure out what is going on but assuming I understand how the 1K resistor fix behaves, I have to guess the trouble happens during the time in which P5 is not driven at all (I'm not even sure this was supposed to be a valid state or a long enough state that the pad would notice).
    The 1K resistor fix forces this floating P5 state to be identical to the 2 button pass thru mode (as if P5 is driven high), once more not sure why it is only necessary on one pad and not the other, maybe the setup time of that particular 74LS125 is different enough to cause the issue, don't know.

    One thing is for sure, using a pin, P6 in specific, as both input and output and alternating it during functioning (by switching P5) is really searching for trouble especially if timing is tight and/or floating states are possible. Commodore should have designed this one much better as Sega did, the dual operation mode [in specific the 2 button only mode] should have been supported internally by the CD32 when it detects a 6 button capable pad and not the other way around, well too late for that, I guess they already had enough messed up logic to handle mouse, light pens, paddles that they ran out of options.
    At least P9 is always an output [either directly BLUE button or a serial output when in 6 buttons mode].

    On a side note the "inductance" on the output [P9] is really just a choke for EMI suppression.
     
    Last edited: Dec 31, 2014
  18. themisterwarren

    themisterwarren Newly Registered

    Joined:
    Jul 23, 2014
    Messages:
    1
    Likes Received:
    0
    Hi signed up just so I could bump this thread... I'm having a similar issue and this seems to be the only forum online on the subject.
    Bear in mind please that I'm a passing Amiga gaming enthusiast and I don't really have any education about electronics... failed physics 12 so there you go :p

    I bought an NTSC CD32 in pretty rough aesthetic shape last April (2013) and have spent the year cleaning it up when I can, also modded Alice with the pin 41 hack so I could play the PAL games without any trouble (excuse me, my pops modded the system... he's the one with the 40-odd years computer experience). Never had any problems with the console itself, or the pack-in Commodore controller. A few months ago I tried plugging a European Sega Mega Drive Arcade Stick into the system and it wasn't recognized by the CD32-100 compilation. Ever since then, I've been having odd issues with the Commodore controller in the 1st controller port, and every game I properly own seems to react differently:

    - first off, I also own one of those terrible MATT Boomerang controllers, and it works perfectly fine in both controller ports, no issues. I also have a mouse that works fine in either port, with or without a controller plugged into the other port.
    - in Morph, only the blue button is recognized, and it quits the game.
    - in Battletoads, only the blue button is recognized and it jumps back and forth between pausing the game and punching. however, it works fine in the second port in two-player mode, and if i swap the boomerang controller in the first port for the commodore one in the second, the commodore controller works perfectly fine in the first port, but the problem comes back after resetting the console.
    - in Chaos Engine, the controller works fine in one player mode until you pause the game. when you unpause it, only the red button works as the action button. in two player mode, same as battletoads, it works in the second port and works fine if you swap the controllers.
    - in Arabian Nights, it works in the first controller port only if i have the mouse plugged into the second port and i jiggle it around and press the buttons on the language select screen, otherwise only the blue button works.
    - no longer have the CD32-100 disc but it was burned at 8x when i did.

    With the limited knowledge I have about electronics, it strikes me that I could have blown a capacitor in the console when I plugged in the Sega stick (it was quite the monstrosity), but that doesn't explain why the Boomerang controller works without any issue, unless it runs at a lower voltage then the Commodore one? I don't know, PLEASE correct me if I'm mistaken so I can educate myself a bit more on the subject.

    Anyway, with this limited information I have from my own experimentation, do you think this could be a hardware fault on the part of the controller, or the console? If it's a controller problem then it's easier to just replace it for me then to get my dad to crack it open and mess with it, and if it's a problem with the console I was going to replace the capacitors and the CD drive anyway in the next couple of months. So I will eventually get around to seeing if a solution will be had from any of this, but like I said, I want to educate myself better with the problem and investigate what happened and why. I can take pictures of the motherboard if need be.

    Thanks for reading and hope to hear some ideas back soon!
     
    Last edited: Jul 23, 2014
  19. retro

    retro Resigned from mod duty 15 March 2018

    Joined:
    Mar 13, 2004
    Messages:
    10,354
    Likes Received:
    823
    The controllers for the Mega Drive are wired completely differently. IIRC, Sega tie all buttons to +5V. Commodore ties them to ground. The sticks add further buttons, of course.

    As such, the logic inside the Sega controller probably won't work, hence the controller won't work. And you could kill the logic. As for the console, well the Sega controllers used to damage a chip inside the Commodore 64, although I seem to recall them kinda working on an Amiga. Maybe check the logic chip (U34) wired to the top half of the control ports, but it's possible you could damage Paula too, I guess.
     
  20. amiga1200

    amiga1200 Dauntless Member

    Joined:
    May 9, 2012
    Messages:
    703
    Likes Received:
    4
    can't say i've encountered the original cd32 adjusted that way (pin 41 for ??? ??? reasons) but i can tell you this for free.
    NEVER EVER plug a controller into a machine it's not intended for.
    years ago, cooked a six button megadrive pad by plugging into a1200.
    computer was fine, pad was fucked. (two wires are to be exchanged on the magadrive pad for it to work on the amiga series, i learned this later as it's required to safely use a sega six button job)
    three button pads work great, but as a practice, don't use anything not intended for CBM.
    can't really offer advice other than, DON'T use six button sega pads, they will die in their default state.
    ...
    check out 'amiga.org' or 'amibay' for some possible help, good luck!
     

Share This Page