IPAC Troubleshooting

[Gatt]

Hi all,

I've finished building my latest iteration and moved into the testing phase.  This iteration is a modular panel.  I'm getting some *extremely* strange behavior and wanted to see what others think.  First, the design...

Design:  Cat6 ethernet cables connected to the IPAC, which pass through a 4-port wall jack with female/female jacks inside, and the controls are connected to Cat6 cables.  Coin/Start buttons are daisy chained grounds, controls are daisy chained until the last ground which is naked wire with the ethernet wire for ground.  No controls have the same ground pattern, one joystick might go to the ethernet ground on up, another on left, another on down.

The problem:
-I've tested 5 joysticks, no direction is recognized except for down on any of the joysticks
-I've tested those joysticks for the ports for Player 1, Player 2, and Player 3, all behave exactly the same
-I've tested the button controls, only button 5 works
-Using a top-fire I've very randomly been able to get button 1 to be recognized (Top-fire button shares button 1), but none of the button controls work
-I've used both leaf switch and micro-switch joysticks with the same behavior
-Coin/Start works only for Player 2
-Note:  I know it isn't a ground issue because some of the joysticks down is the last in the daisy chain, so the others must ground.  All of the coin/start buttons daisy chain ground with Player 4 running to the IPAC, so at least Players 2-4 must be grounded.
-Note:  I've pulled some of the wires and reinserted just to make sure it wasn't bad connection.

I suspect that my IPAC4 is bad, but wanted to get other takes before I bought another.  One disclaimer, my puppy got ahold of the IPAC4 awhile ago, but it didn't look like he really chewed it, looked like he was more interested in the box.  But I'm suspecting it's damaged now.  The only other thing I could think of, which feels strange to me, is that Cat6 can't really carry the electricity to the IPAC because the wiring gauge is so small.  I've used Cat5 in the past, but Cat6 seems thinner to me.

So...thoughts? 

[Laythe]

I'd take a jumper wire and sit there test-jumping a GND pin to an input pin on the IPAC itself, with something like Notepad open.

Ignoring all the real controls, cat5, etc - does grounding the input pin make the appropriate keystroke output happen? 
  If so, the problem is downstream of the IPAC, in your cat6/jack/controls/etc.
  If no, the problem is in the IPAC, you might try reflashing the program on it, or, it may be damaged.

[Gatt]

Thanks!  When you say Ground Pin, do you mean touch the terminal screw or running a wire from ground to say "Up" for Player 1 and clamp both terminal screws down on it?

[Laythe]

Oops, sorry, I had the wrong ultimarc board in mind.  I see now, the ipac4 has screw terminals, not pins.


The switches in your joysticks and buttons just make a connection between two wires.  One wire is normally a daisy chain from one of the two screw terminals marked GND, and the other wire is connected to the screw terminal of that specific named input.  What the ipac4 is detecting is simply continuity between GND and the selected input.

I think the outside of the terminal screws - particularly when screwed in tight - also have continuity. 

So, yes, it'd just be a jumper wire held with one end touching the GND terminal screw and the other end touching the screw for, say, "up" for player 1.  If the wire touches both screws, the ipac ought to see that control as activated and send the corresponding keystroke to the computer.  Running notepad or something will let you see the keystrokes getting typed.

This is just a method for bypassing everything besides the ipac. 

If a jumper wire touching GND and a specific control screw terminal doesn't work anywhere, then I'm wrong and the screws can't be used that way.

If a jumper wire touching GND and a specific control screw terminal only works for the controls which you noticed still work, then the problem is inside the ipac.  It's not necessarily fried, it might just have a partially empty program loaded.  Or it might be fried.

If a jumper wire touching GND and a specific control screw terminal works on all the controls, then it's something in your controls or wiring, the board is working correctly and you can move on to diagnosing your controls.

Hope this helps!

[Mike A]

A picture of your wiring job to the ipac would be useful.

[Gatt]

I followed Laythe's instructions, when I closed the loop each of Player 1's controls worked.  So it must be something in the path between the IPAC and controls.

Thoughts about CAT6 being too thin?  It's nearly impossible to strip it down without losing some copper filaments, making an already thin wire even more thin.  Given that it's present in each of the 6 keystone Female/Female keystone jacks and consistent across controls each of which have different wires, I'm wondering if it's a wire gauge problem now.

When I wired in to the IPAC I actually stripped the wires long and folded them because they're so thin, so I'm pretty sure they have contact.

It just seems strange to me that it's consistent across many controls and different players.  Player 1, Player 2, and Player 3 are all on different keystone jacks with different ethernet cables behind them, but all of them have the same problem, only down works for 5 different joysticks.

[Gatt]

I'm really really stumped now and wanted to reach out to see if anyone had any ideas.

I got to dig deeper into this today, I tried swapping the wiring in between the controls and IPAC and came up with the same result.  I then pulled out an old IPAC2 board, and it's giving me exactly the same behavior.  I checked the mappings on the IPACs and the mappings match MAME.  I used notepad and the only character that registers for Player 2 is 'g' across several different joysticks.  I inspected the control wiring and can't find anything off. 

The only thing I can think of at this point is something weird in software, particularly Windows.  Anyone have any other ideas for me to try?

[Laythe]

I think your wiring and/or controls are the suspect, if using a wire jumper on the ipac between ground and each input produces all the correct keystroke events in notepad.

You might try, as a test, wiring up one simple button on a new heavier wire for just one of the inputs that isn't working, like p2 right.  If that works, see if it still works on cat6 wire.  Swapping out things like that and doing methodical science to it should help narrow down what it is. 

You might also try unscrewing all the wires off the ipac on your existing wiring harness, and connecting things one at a time to see if they work, or if at some point adding one thing takes a lot of stuff offline. 

[Gatt]

Thanks Laythe,

I figured it out!

Summary:  The female to female keystone jacks are reversing the wiring order.

Story:  I tried one last joystick, my Happs flightstick.  Saw the same behavior, except hit one bizarre thing.  Button 2 was registering as "Up".  Which was incredibly bizarre, because I only wired joystick directions to the test IPAC2 board I pulled out of the closet.  I suspected the female to female keystones were reversing wiring order between them after looking at the pinout for Cat6, and dismantled one of the keystones.

Sure enough, the little PCB on the keystone has numbers showing wiring order, they're opposite order between the two ends of the little PCB.  So the joystick is sending a signal on say Orange, the keystone is sending it back out on Brown, which isn't connected.  "Down" works because it's in the middle of the swapped wires, the other three are sending signals on wires that aren't connected.

Thank you again for your help!

[Laythe]

Aha!

That's sneaky.  Good job figuring that one out.   

Sure thing, glad to be of assistance.