Button press wiring question

[demogo]

LOL!

I just went to order some parts from Bob Roberts for my kit and I told him what I was planning on doing and when I mentioned that I was getting a 'ground' from the other button wiring he said the same thing I was told here earlier -- you flat out have a wiring issue somewhere in your cab.

He said that floating ground or not, the only thing that would cause the button to act as if it had been pressed was if the other wire was in fact a ground.

So crap, what to do now? I don't want to hack everything up just to find out that I had a wiring issue all along that was causing these problems!

Ken, I just checked the ground on the screw of my inactive GL board against the cabinet ground and it was definitely grounded perfectly.

I just received some jamma extension cables from jammaboards and interestingly enough when I set the DMM to 2000K ohm and touched the ground wire (pin 1) on the jamma extension harness running the (active) NBA Jam board I got a reading of about 560. Shouldn't this have been 0 or close to it?

Also interesting was that I got no reading at all from pin 28 on the jamma extender harness and I'd think that'd be at the same reading as pin 1?

Am I missing something here?

I'm gonna go move the extender harness to GL now and probe those same pins...

 

[demogo]

I got negative numbers on both probes of the jamma extender cable when GL was active... so my guess is that the NBA Jam just didn't route anything to that pin when it was active?

But neither probes resulted in numbers like I expected either.

Can anyone give me an idea as to where to start probing with my DMM? I'm not even sure where to begin...

Not only do I have this unexplained grounding effect I also have what appears to be hum bars or some kind of video noise in my Jam picture only. When GL is active the picture is perfect, no noise. When Jam is active I get noticeable interference lines that run through the picture. When I tried this same Jam board at someone else's house it worked w/o any noise.

Could they be related? Does anyone have any ideas as to where I could start?

Thanks in advance!

 

[demogo]

Anyone? Any comments?

Can anyone at least tell me what kind of values I should be getting back from the buttons/joystick wires from my unpowered GL?

What does success look like? What should I be seeing?

 

[cdjump]

it could easily be that their is enough flow from the unconnected boardset through to ground that it could cause your powered board to think it was pressed. A diode isolates it and keeps it current from flowing backwards through the other boardset.
As far as the interference- it could be where the board is mounted - like too close to the power supply.
As far as your wiring goes - put your dmm on diode test, and then probe the button in question, with the other lead to ground. You should get open on one of the button terminals, and ground on the other. If you don't read open, chances are your seeing backflow through the other boardset. Putting a diode on it will correct the problem.
Now, on checking your voltage- You should check it on the boardset itself - preferably on an eprom or logic chip (bottom left corner, and top right corner, with the notch facing up) It should read about 5.1v. The signal voltage at the button is meaningless, and could fluctuate wildly below what the board itself is seeing.

 

[demogo]

Well, it would be good if it were as simple as what you're describing and not something that would make me rewire the whole cab.

So moving the board to try to minimize the interference didn't help -- it actually seemed to make it worse.

You can see the interference lines scrolling through the picture and then actually see them slow down, stop, and then start rolling through the picture in the other direction. At times they are visible, at other times they disappear.

On the wiring, I *think* I did what you suggested.

I grabbed a button, hooked up the ground from the *active* game (NBA Jam) to the ground prong of the button, then hooked up the connector for that button from the inactive game (GL) to the NO tab. Then I took a wire from the active game (Jam) for the same button and touched it to the NO terminal where the GL button was hooked to. The game saw the button press (which is the bad behavior) and then I set the DMM to diode and put the DMM black probe on the *active* game's ground and the red probe on the NO terminal (which still had both games wires on it) and got '1' which is open I guess. When I moved the red probe to the NC terminal I got a low number that was between 010-020 (which I assume is ground).

Is this what you had in mind for testing?

If so, does that mean that I'm NOT seeing backflow? Or did I have my ground probes in the wrong places? Or messed something else up?

Haven't had a chance to check the voltage on the game board yet as you described but when I first adjusted the power supply down to 5v I had pulled a connector to the sound board and check the +5v line on the connector and it was right at 5v. I will check as you described tomorrow as I'm now very late to bed!

Thanks -- I'm very appreciative of the help!

 

[cdjump]

thats basicly what i meant. I wouldn't worry that your cabinet has problems if the other game plays fine in it. The issue is somewhere in how the 2 games are connected together. The only thing is when you use the diode test, you shouldn't have power applied to the circuit.
One more thing to try. With the game powered on set your voltmeter to +20vdc, touch the ground to the ground, and the red lead to the NO on the button -with it connected to the active game. Then hook the switch lead from the inactive game to the NO terminal as well. Does the voltage drop? If it does, your getting enough current flowing backwards through the inactive board to cause the active board to see a low logic level and interpret it as a button press.
You can also try adding a diode to the switch grounds after they split apart and go to each board, with the band facing each circuit board. That should stop current from flowing backwards, and only require 2 diodes, instead of 28 or however many switches there are.
One more thing to try on your interference. Turn the power off, and unplug the game. Put your meter on continuity test (or diode test), and test continuity between logic ground, and earth ground. If there's not any continuity, run a jumper wire between the two. I've had similar interference issues on 2 different games, and connecting earth ground and the logic ground solved it both times.
And also - checking the voltage on at least one, if not several chips on the boardset itself is important. Its not unusual for there to be a good deal of drop from one board in a stack to another, or even from one side of the board to the other. Just because you've got 5v where it goes onto the board, isn't a guarantee that you've got it throughout the board. This is why you see lots of games that run fine at 5.2 but not at 5.0- the whole boardset isn't seeing the 5v its needs (or 4.8).

 

[demogo]

Thanks.

Yes, GL plays perfectly in the cabinet.

Per your comment about the interference:

Is earth ground in this case the cluster of ground wires that go to the frame of the cabinet? And logic ground a ground point on the active game somewhere like where a screw would secure the board?

 

[demogo]

> The only thing is when you use the diode test, you shouldn't have power applied to the circuit.

I assume you mean that both games should have been turned off?

If so, I just repeated the same test that way and the results are different.

I get a ground from the NC tab of the button. From the NO tab I get a value of about 1900 with just the "active" Jam game (hooked up to the Jamma cable). When I added the wire from the "inactive" GL game the value on the NO tab dropped to about 1300.

 

[cdjump]

that means that it takes 1.3v to positive bias the inactive game, and allow current to flow. The problem is that the voltage to the sense lines is usually on a large value resistor, so very little current actually flows. When you allow it any path, since the resistor is so large, it drops the voltage drastically, which the game sees as a button press.
I assume that all of your buttons are run on a common loop that grounds back to the board. All you need to do, is break that ground, and add a diode, with the band facing the circuit board. Then you have another ground that goes to the second board's switch inputs. Break that ground and put a diode in with the band facing the circuit board. That allows current to only flow through one way, and stops the reverse current your getting.
One last thing on the diodess- a 4000 series diode is fine, but overkill. A 1n914/4148 diode is all you need - crap shack sells a 50 pack of them for about 3 bucks.

And, i assume you have either a switcher or atx power supply. Logic ground would be any ground from the circuit board, or the ground coming out of an atx power supply (black wire). Earth ground is the green wire that comes in from the power cord. If the ground prong's broke or not hooked up it could cause that as well. If you have a switching power supply, it'll have all of the connections right there together, and you can just probe them there, and add a jumper there. It depends on the brand of the switcher - some say gnd and earth, and others say different things. like dc ground, ac ground, etc.

 

[demogo]

Thanks for all the feedback, cdjump

I believe I understand your points with the wiring now.

At this point though I just have the one ground loop -- I don't have separate ones for each game.

The stock GL wiring brings a ground (and +5v) to each player control (to power the 49 way sticks) but this isn't going to work if I'm going to share the controls with NBA Jam.

So my plan is to take the ground that comes up to the player 1 controls from the Jamma harness and 'jump it' over to p2, p3, and p4. That way I can use the new 8 way sticks with both Jam and GL. At the moment the only player control that is hooked up is P1 and it has an 8 way stick hooked up with 4 buttons. All of the other testing I've been doing has been directly connecting the wires and using the p1 ground signal (which I'm assuming is part of the jamma wiring).

But that just leaves me with one shared ground loop for both boards at the same time and I'm not sure how I can use your diode trick with the ground to help with that.

Having said that, it certainly seems to make sense to use a few diodes on the ground rather than on 20 something controls.

As far as connecting earth ground to logic ground I'm a little lost. I tried connecting a black ground wire on the CP to a green/yellow striped control panel ground wire and that didn't help.

This is what my power supply looks like (see attachment). I'm not seeing anything obvious here that I should be hooking up to. All of the output comes out in a series of molex type plugs that are described on the label.

I'm not sure what to do next with the ground...

P.S. I just checked continuity on the ground plug from where it plugs into the back of the power supply all the way to the electrical socket where it plugs in and continuity was good so I think I'm okay there.

 

[TK_Monkey]

One last thing on the diodess- a 4000 series diode is fine, but overkill. A 1n914/4148 diode is all you need - crap shack sells a 50 pack of them for about 3 bucks.

Are those specific to his ground situation or would they also work for combined wiring of kick harnesses on the jammaboards 6-1 which in theory only should be powering one board at a time? If the end result is the same, I'd much rather spend 3 bucks than 50.

 

[demogo]

Measured the voltage on 2 ROMs (upper right side near the notch) at 5.12 volts.

Video noise is definitely still there. It seems worse with the Jammaboards extension harnesses though.

Having other issues with the extension harnesses as well so I guess they're going back tomorrow and I'll try a $$$$ Bob Roberts version. Bob's cost more than 2 from Jammaboards combined but I'm betting they'll work much better.

 

[demogo]

OK, new post to an old thread and I'm hoping someone will notice and respond.

I've had success using diodes to prevent interference between a control being hooked up to 2 different boards at the same time.

Now here's the wrinkle and I'm hoping someone can answer my question.

Diodes work great when only one board of the 2 is powered at a time.

What happens if you are running a jamma switcher that powers both boards at the same time so that both diodes which join together are both powered? Will the diodes still prevent interference between the two games? What will happen when a button is pressed?

Thanks in advance!

 

[cdjump]

nothing should change. Both boardsets will see the button press, nothing more. (I'm assuming your talking about a kick harness) While both boards are powered, only the one selected should see the coin up inputs. It'd be just like walking by a machine sitting in attract mode and pressing those buttons. In most cases, nothing happens. Worst case scenario is the your actually playing both games at once - if the unselected game managed to somehow get coined up.

 

[demogo]

Excellent, thanks!

It would actually be buttons on the jamma harness that would be involved, not a kick harness.

Now that I've got everything working well on my GL / NBA Jam cabinet, I figured the next step is to go the jamma switcher route and I'm leaning towards Mike's Arcades 2-1 jamma switcher.

I've been told that it powers both games at the same time and only sound and video are 'switched', everything else passes through to both games. Kind of weird, but if what I was told was true then a coin drop would be seen by both games and you'd literally be playing two games at the same time, even though you couldn't see or hear the 'other' game.

I'm not thrilled about having both games up all the time but I basically have a choice between Mike's switcher or Clay's and they both act the same in this regard I believe.

The only other switcher that I've heard of anyone have success with was the 6-1 that jammaboards sold and you had to hack it to get it to work. And not only is it not available anymore but I've just ran into someone else that said it didn't work with a crap with his games.