[Solved] Another 6 buttons issue

[Takosuke]

Heya,

I've been reading most of the threads about 6 buttons issues and tried a lot of things without success, so coming for help.
Bought this week the Arpicade full pack with kick harness from Highscoresave (Horizontal 4000+ games).

My cab is a Kyotaro 29" cabled for CPS2 (so the P1 4,5,6 and P2 4,5,6 are cabled with a CPS2 Harness).
I used the Highscoresaves harness and put cables it to the 2nd pin of the 4,5,6 buttons, and grounded them as well back to the regular loop.
plugged everything to the pi board, and set the dipswitches to 1&2 on, 3&4 off so it uses regular Jamma (which is what I have since the cab was configured for CPS2).
Did not modify any file since the fresh install.

Here are the issues:

• FBA, PS1, SNES games: Everything works fine except button 6
• Mame172 games (ie SF Alpha series, Marvel vs Capcom etc) the controls are all weirdly working: the P1 controls act as P2 controls (P1 starts makes P2 entering the game, P1 joystick and buttons works for P2), P2 controls are not responsive at all, and off the 6 buttons,only 5 works - 6th button does nothing. I've checked the setting by pressing tab, and going up does not work, but accessing the button assignement, everything looks correct (P1 up is Joy1 up etc)
• Advmame078 games like MK2, 3, 3U: P5 button is stuck in always pressed (makes the character block all the time), if I keep P1 or P2 button 5 pressed then it allows me to press other butrtons

Does anybody have any idea? I tried moving the dips to Jamma 6 (1&2 off and 3&4 on) and same result. if I remove the harness, then I only have 3 buttons in this config.
I tried different ground point for 4,5,6 and same result. If they are not grounded, they are not working, as it should be, so I am sure they are grounded.

I've ordered JNX Atlas converted to try to see if plugging my CPS2 Kick Harness in it and setting it as Jamma 6 would work as I am out of ideas

I can post pictures if needed (let me know what pictures to take).

Thank you!

[dee2eR]

That's certainly odd. I don't know if this one will be a hardware issue (maybe the b6 not working issue) as that shouldn't/couldn't cause inconsistent behavior in the different emulators.

I'd recommend downloading the vanilla version of ARpiCADE 3.83 to test. Using a vanilla release at least should remove the possibility of software setup problems (I have no way to know what has been done on the build you bought).

To test inputs I generally use Street Fighter 2 Hyper Fighting in MAME0172 as it's test menu shows all inputs (except service and test).

You should not need the JNX adaptor to use 6 buttons with the RaspberryJAMMA but I can say from personal experience that it will work with the RJ in 6 button on JAMMA mode > JNX board > CPS2 kick harness. If you haven't already ordered the JNX maybe hold off ordering for another day or so as you should be able to set it up without it.

Lastly, if you have the high power wiring for your monitor and the button signals all bundled together you may be experiencing interference from the power wiring - I haven't had this issue myself but it has been reported a couple of times now. If you wiring is all bundled together like this you may just need to separate the 110v for the monitor from the button signals.

[Takosuke]

Thanks! Just installed the 3.83 image. Launched it with just the SF2 hyper fighting rom as the only modification after and tested it right away.

Same behavior:

I can add credit, but My P1 controls work for P2 and B6 does not do anything, if I try to access the service mode I can't do anything or select an entry in the menu as the P1 controls do not respond (looks like it's stuck going up/left though as the menu keeps scrolling).

Note that if I unplug the kick harness from the pi board, it does exactly the same (except obviously not having b4 and 5 working this time).

If I remove the Arpicade board and plug my CPS2 A+B Boards and the cps2 kick harness, no issue at all and 6 buttons work fine.

Also checked, and the 110v cable for the monitor is far away from the button cables

[Takosuke]

Got at least something working

Turns out that having my wireless keyboard plugged in on the Pi created the weird issue with Mame, making all the controls messy (P2 registered as P1 and P1 always going up/left).

Unplugging it make it so P1 and P2 behave normally. Now, 1 button is still somewhat not working when the harness is plugged in (no high/roundhouse kick on SF2 HF for P1, but for P2 it seems that B3 is not working, ie hort kick).
According to to the emu controls screen, ti would translate to Joy1 button 4 and Joy 2 button 3. If I try to manually assign them, they don't register.

[dee2eR]

That's interesting, I wonder if you wireless keyboard has a gamepad emulation mode or something, it sounds like it was becoming joy 1 (joy0 as the kernel sees it) making all the settings wrong. You may find a different keyboard will not do this, I haven't seen that happen before.

B3 should be strong punch in Street Fighter.

Are all your buttons using a daisy chained GND? There's not a couple of different GNDs that aren't connected with the rest or something?

Is there any visible damage on the ribbon cable connecting the RaspberryJAMMA and RPi?

Do you have a second Pi you could try with the RaspberryJAMMA to rule out the Pi as the source of the issue?

Unless I'm missing something the only other physical fault that could cause the issue is a short or a (couple of) bad resistor(s) on the RaspberryJAMMA. I test all boards before sending them out so I would be very surprised of a resistor went bad for no reason since testing but it could cause the issue. It's worth having a look for a short, maybe a wire is bridging some inputs to GND which would stop them working too.

[Takosuke]

Thanks for taking the time to answer me! I appreciate the help there.

I borrowed a regular USB keyboard from work and don't have the issue with it, so your assumption is probably right for the mini wireless, maybe the mouse emulation on it or extra functions are messing things up (here is the model for reference https://www.amazon.com/Upgraded-Aerb-Wi ... B07JH8DZ4R).

I did test the continuity of the ground daisy chain, and it seems to work. I did a crappy test though and should receive my new multimeter in a couple of days to make 100% sure there is no issue there. But if I don't touch the daisy chain and change the P1 B6 cable to P1 B5, then the B6 button works as B5 and B5 stops working. Similar results if I switch the cables between P2 B4 and P2 B6.
That should eliminate the doubt on the ground and also a faulty switch on the button.

(the layout of my panel is 3 buttons on top: 1,2,3, 3 buttons on the bottom: 4,5,6)

P1 B6 and P2 B4 are not registered in the service mode IO test of SF2HF (tried to move it to the latest mame and similar result)

I am aware of the weird naming of each inputs, but also did a jstest /dev/input/js0 and nothing registers if I I press P1 B6 (nothing happens).

I don't see any damage on the ribbon, although there is a microscopic peck of plastic a little brushed off on the wire at the edge of it (complete opposite of the red one) but it seems very superficial and does not reveal the wire under. Could that be the culprit?

All the pins seem ok at a glimpse, and I don't see any damage on the circuit board either. I could check it more in depth once I receive my new multimeter.

I do not have an extra pi to test, unfortunately. I can see if one of my coworkers can lend me one. I currently have a 3B+ on it, would a 3 work the same?

[dee2eR]

Well it sounds like everything is OK on the cabinet side. That at least rules that out. And if the buttons are not registering at all and not stuck in an ON state that should rule out a short.

Do you have B4 connected through the JAMMA edge or the kick harness? If you haven't already it's worth checking the kick harness too, especially if both the problem inputs are on the kick harness. Check you have continuity from the pins inside the connector at the RaspberryJAMMA end all the way to the button(s) just to rule out one or more poorly contacting crimps or broken wires.

After that the only things left that it could be are dead resistor(s) or broken tracks on the RaspberryJAMMA, ribbon cable damage, or an internal problem in your RPis GPIO.

If you can try another RPi, the Pi2, Pi3, Pi3B+ and Pi3A+ on will work with the current ARpiCADE, Pi2 shows some boot text but will work otherwise.

Player 2 B4 is either the last or second last wire of the ribbon cable (sorry can't remember which offhand, but it is definitely one of the last two on the opposite side to the red line) so it is possible the ribbon cable is the cause. If you have a 40pin IDE cable around you may be able to test with it, provided it is fully populated and not keyed with a blocked hole or something. Alternately you could poke something into the ribbon cable at each end to test it's continuity (the little 2 pin joiner included with the RaspberryJAMMA for the reset header would probably work if you didn't chuck it when you installed your Pi), or maybe your meter has probes fine enough to do it. P1B6 isn't that close to the end of the ribbon cable but who knows...

Lastly if you want to trace out the two tracks on the bottom of the RaspberryJAMMA board they're both located right next to each other on the kick harness header (the middle two pins) so I'd start from there. There's only tracks and resistors so it's an easy job with a multimeter.

[Takosuke]

I don't have the B4 connected through JAMMA, but the kick harness.

Tomorrow I will get a couple RPis and test with them, and also check all the connections with a multimeter and report back. I did not install the pi as I took the full kit from High Score Saves, Harness, PI and preloaded SD card included (now updated to 3.83 Vanilla ).

I'll let you know after tomorrow, thanks again for the follow-up!

[Takosuke]

Heya!

So I finally found where the issue was, and it was an anticlimax

The harness I bought from High Score Saves had bad soldering of the wires for P1 B6 and P2 B4 on the connector that plugs into the board.

Once I received my new multimeter (the old one was DOA), I tested everything for continuity and voila, the issue was isolated. I resoldered those wires properly, and now everything is working flawlessly

Thanks a lot for the support, and at least I've learned through the process that my wireless mini keyboard was no good, and got a promo mini usb keyboard now I'm editing the post to add that the issue is resolved!

[fuzzbee]

Question - Is it possible to wire button 5 and 6 on the Jamma edge of the RaspberryJamma Board or do I HAVE to use the kick harness and wire to the 4 pins on the board?