Severe Ghosting problem with J-PAC

[BuriedEgg]

Hello,

I've been lurking for a long time but this is my first post.

I've had a problem with certain keys auto-triggering themselves for a few months now and I can't figure it out. First I will list my setup:

*MAME PC running Atomic FE.
*Windows XP.
*J-PAC with latest Winipac for config.
*Converted 1980's Dig-Dug cabinet to MK1 cabinet.
*6-buttons/player - working Coin Slots and Test Mode Switch.
*Original MK1 JAMMA+ harness with kick harness for extra buttons all wired directly to J-PAC via edge connector and screw connections.
*LCD monitor 3:4 ratio connected directly to PC (video NOT thru J-PAC).

The problem is that any button connected to the screw connectors (P1&P2 --BINGO! Either that, or I was attempting to say "before" but it was too many letters to type--, B5, B6) will automatically trigger itself at random. Also, F1, F2, & F3 will auto-trigger. This means that I'll be right in the middle of a game and if F2 auto-triggers it will dump me into Test Mode and end my game. It is important to note that the autotriggering can sometimes take 10 minutes or as much as an hour or two to happen. It is intermittant with no discernable pattern.

To describe the troubleshooting steps I will use P2-B6 as an example. P2-B6 is mapped to letter "V" in Winipac:

I thought it was bad wiring or a short in the old MK1 harness so I disconnected everything from the J-PAC. All buttons, joysticks, coin and test switches. Then I used two wires to connect directly from a pushbutton to P2-B6 on the J-PAC screw connectors. It still auto triggered. I opened a text window in Windows and I could watch the letter "V" randomly typing itself across the screen.

I thought maybe it was a bad microswitch so I tried connecting a different pushbutton to P2-B6 on the J-PAC and it too did the same thing. Letter "V" randomly typing across the screen. Remember, this is without anything else connected to the J-PAC. The J-PAC is connected to the PC via the USB cable that came with it.

I tried disconnecting the extra USB keyboard from the PC and it still did it.

I tried connecting the J-PAC with all the buttons to a different PC (my laptop) and it still did it so that eliminates my MAME PC as the problem.

When the main JAMMA edge connector is disconnected from the J-PAC the F1, F2, and F3 keys do NOT autotrigger. When just one button is connected to the extra screw connecters, with nothing else connected, then only that button will autotrigger.

We've tested the wiring with a multimeter and continuity of the solder connections and the wires are good.

Andy of Ultimarc was kind enough to send me a replacement J-PAC and it still auto-triggered the same buttons.

I have included a video of the troubleshooting steps and what my setup looks like to aid anybody who wants to help me figure this out.

Here is a link to my Photobucket page. Its the first two videos on the top-left labelled "Malfunctioning". They are identical. One is LowRes and one is HiRes:
http://s2.photobucket.com/albums/y45/Dickjones/Arcade%20Cabinet%20Project/

(NOTE: In this video I used spare wires connected with wire-nuts spliced in in
the middle. After filming I also tried direct wires with no splices and had the
same results.)


Any advice would be great at this point as I am out of my mind with frustration. This cabinet was working for about 6 months and then this started happenning. I don't recall making any changes to it at that time.

Don't be shy about asking me to provide more photos or information or to try something. I'll do almost anything at this point.

Thanks in advance for any assistance or insight anyone can provide!

[BuriedEgg]

Thanks for looking at this with me!

-J-PAC is secured. Unable to come into contact with anything inside cabinet.

-Tried PS2 adapter. It still auto-triggers the buttons. Seemed to increase the frequency of auto-triggers but that may be my own subjective view.

-When J-PAC is connected to PC without any inputs (buttons, JAMMA harness, Joysticks) it does NOT auto-trigger any buttons.

If just one button is connected to P2-B6 screw connector it will auto-trigger. I've tried differnent physical buttons, different physical microswitches, even mapped different letters to P2-B6.

P2-B6 is the main culprit but, buttons 4 and 5 do it too. Just not as frequently. Is there any relevence that it is only the screw connector buttons that do it?

F1, F2, and F3 do it but not nearly as frequently. They are not connected to the screw connectors.

[BobA]

Could it be that your jamma power supply is not stable or set at 5V?   A flaky PS can cause alot of problems.

Check your voltages.

[BuriedEgg]

That is an excellent question. I'm glad you asked. I did some testing with a multimeter. Maybe you can tell me if these ranges I'm getting are okay:

At the link below you will find 3 videos of me testing with a multimeter. One is a video of testing right at the screw connectors with the entire MAME cabinet
powered on. The other 2 videos are of testing with the power off at the screw connectors and then again right at the pushbutton microswitch connectors (Best
to test at both ends of the wires!).

Here are the videos of the tests/results:
http://s2.photobucket.com/albums/y45/Dickjones/Arcade%20Cabinet%20Malfunction/

Here are the results:
-Testing with PC powered on it is 4.92 when the button is not pressed. Is 4.92 sufficient? I thought it was, but you may have a lot more insight into correct voltages than I.
-Testing with PC powered off (at the screw connectors) it is 0.04 when button is not pressed. It is 0.00 when button is pressed.
-Testing with PC powered off (at the pushbutton microswitch connectors) it is 0.03 when button is not pressed. It is 0.00 when button is pressed.


Finally, here is something that might help. I connected the multimeter wires directly to the P2-B6 pushbutton connectors while the MAME cabinet was turned on. This gave a reading of 4.92 when the pushbutton was NOT being pressed down. I tested pressing the button down and the voltage dropped sharply and quickly each time the button is pressed, as it should. So, I left a text window open in Windows and I left a video camera recording the multimeter for one half hour. When I came back I saw 5 letter "V"s had automatically typed across the screen (Letter "V" is mapped to the test button P2-B6). When I viewed the half hour of video there was no change in the voltage. The 4.92 stayed the same (4.91 - 4.92) the entire time!

So, When the P2-B6 key was being activated randomly the voltage at the microswitch (the physical pushbutton) never changed. This tells me the false signal is not coming from my wiring or my pushbuttons. It must be happening at the J-PAC or the PC. I've tried multiple J-PACs and two diffent PCs (Windows XP and Windows 7) and it still does the auto-triggering.

[BuriedEgg]

The only power supply I have in there is the PC's own power supply. It powers the PC. I do not have an arcade cabinet/JAMMA power supply.

I have tried different USB ports. Actually, I just realized that when I tested the PS2 adapter on the J-PAC cable I forgot to set it to "PS2" in the Winipac software! I'll test again tonight.

I should also point out that I have tested 3 different J-PACs and they all auto-trigger those same buttons. Andy at Ultimarc has been really terrific in supporting me on troubleshooting this. He sent me the extra J-PACs to test with. I don't know how they could all have failing PCBs. That was my guess as well, at first!

When I test the screw connectors tonight, can you tell me exactly what you'd like me to test? Do you want there to be a load on them (PC power on) or not when I test?

Thanks for the recommendation on that KeyWiz!

[BuriedEgg]

Okay, so I performed all the tests listed above. Thank you, everybody, for the helpful suggestions. Here are the results:

-Tested a PS2 connection (by attaching a PS2 adapter to the USB cable provided by Ultimarc).

-Have not found a shorter USB cable than what came with the product from Ultimarc.

-Disabled USB power-saving.

-Installed fresh copy of latest Winipac on a different PC (Windows 7) in the house and reprogrammed J-PAC on that PC.

-Disconnected J-PAC from arcade cabinet completely. Grabbed a spare unused Happ pushbutton, spare unused microswitch, and fresh unused wiring. Connected them alone to the J-PAC. Just one button connected to P2-B6 screw connector on J-PAC. Connected J-PAC via USB to brand new Dell StudioXPS laptop from work (Windows 7). Moved entire setup to another floor in the house far away from any electronics/speakers/cellphones/radios.

Through all of these tests it still auto-triggered every time. The most disturbing one was when the laptop was auto-triggering all by itself. I mean its a completely different PC, button, Wiring, OS, etc.

I've tried:
-Different J-PACs
-Different PCs
-Different OSs
-Different Winipac versions (up to latest posted on website)
-Different buttons/microswitches/wiring.
-Different button mappings (different letters mapped to different keys)

I can't isolate it. I don't know what to do except maybe try a different brand of key encoder. KeyWiz?

I think I need to state, at this point, that Andy at Ultimarc has been absolutely outstanding with his support for this product. He has been in constant contact with me over email for the last couple months now trying everything he can to troubelshoot this thing. He even sent me a couple J-PACs free of charge, in good faith, to test out. I've never seen customer support like this on any product I've ever purchased so I hope my experience hasn't put anybody off of Ultimarc's awesome products and support service.

I just don't know what is so unique about my situation that this particular product will not work on any application in my house.

I have a video of the laptop test I performed if anybody wants to see how I've wired it. Its the first video on the page (below) labelled "Laptop J-PAC Test":
http://s2.photobucket.com/albums/y45/Dickjones/Arcade%20Cabinet%20Malfunction/

Thanks for the help and any further insight you may have!

[Gray_Area]

....what about MAME controls configuration, I-PAC shift function, and sticky keys?

If none of those is the issue, I can only guess you have an arcade ghost. In which case, you'll have to find the secret power pellet, hiding somewhere in your home.

[Forces21]

I noticed this as part of your test procedure

-Tested a PS2 connection (by attaching a PS2 adapter to the USB cable provided by Ultimarc).

Have you tried this with a dedicated ps2 cable? not an adapter? this would be a ps2 male to ps2 male cable.
I'm just asking but I assume that although you have tested accross 3 different JPAC's you have never changed this cable? Maybe this could be the problem?

Just putting out some more thoughts for you.

Regards

Forces21

[Lilwolf]

When you tested the other buttons, I'm assuming you did it with new cable right?  (ie, you didn't just plug one in that you already had wired, then another)...

If so, it sounds like you tried everything.  You should really contact ultimarc.  He's really good at helping out and might have a suggestion or two.

[BuriedEgg]

Yes, I used new cable to wire each button when testing. I tried using a new microswitch, and brand new pushbuttons from Happ as well.

I will also try to find a dedicated PS2-to-PS2 cable. I have not tried that yet. Only tried with PS2 adapter so far. That is a very good suggestion.

What about "sticky-keys"? You want me to disable it? I'll check if its enabled. I don't think that's even enabled by default is it (Windows XP/Windows 7)? The second PC I tested on was a brand new Dell XPS. I'll check anyway.

Remember guys, nothing is too silly to suggest to me. All of the above advice is valuable to me. A lot of the time the strangest thing will usually be the thing that turns out to be the problem all along.

I'm tearing my hair out here!

Gray_Area: I will search around for that power pellet. I will also setup a camera in my gameroom at night and put talcom powder on the floor and get a Ouija Board going!

Remember, nothing is too "out there" to suggest at this point.

[BobA]

+1 for pinballjim - Isolate your jpac from any interference.   Since it only gives key strokes when wires are connected then maybe the wiring is acting as rf conductors (antenna).   Even better than an antistatic bag is good old metal.   Place the jpac and button wire in pie tin (be sure to insulate with a sheet of plastic) I use freezer bags for temp insulation.  Cover the pie tin or baking pan with tin foil.  If possible ground the metal to a 3rd wire ground or a metal pipe if you have metal pipes in your plumbing. If you have newer plastic piping this will not work.  This basically creates a faraday cage around your jpac and button wire.

If this works then bring your button wire outside the faraday cage to see if it will auto trigger.   If it does you can try running your wire thru a ferrite core to get rid of the triggering.   A couple of turns around a ferrite donut would be sufficient.   

This is just testing to determine if radio frequency energy could be the cause.   If it is then the source of the rf would have to be determined.  Are you close to any of the following?   High Voltage transmission lines,  Microwave relay tower, Cell phone tower, Radio or TV transmitter tower

[BuriedEgg]

Oh no! I performed the following test:

-Take one fresh wire and connect it only to the positive screw connector for P2-B6.
-Do not connect button, microswitch or Ground wire. Just one wire connected to J-PAC screw connector P2-B6 and let it hang loose at the other end.
-Connect J-PAC to laptop and move laptop to a totally different room far away from Arcade Cabinet.
-Leave text window open for 20 minutes.
Result: It typed the letter "V" all across the screen. The wire had no ground! It was acting like an antenna!

Next:
-Take J-PAC and wire and bag them in plastic sandwich bags and wrap them in a couple layers of tin foil.
Result: With the tin foil wrapped around it, it did NOT type the letter "V" anymore. Its been about 20 minutes. I'll keep testing but I'm scared about bigger things than arcade games now.

This (See Attached Image) is behind my house less than 25 feet away in our backyard!

We just had our first child. He's only 2 months old. I need to contact the County first thing in the morning. Maybe the local news station too. In the meantime, is the KeyWiz less susceptible to signal noise from things like power lines? Should I still purchase one, or just focus on getting the county to tear down those lines in my backyard?


ADDITIONAL INFORMATION:
Its been about 15 minutes since I posted the information above and I just checked it again and it indeed, HAS typed a random letter "V". I will take the whole setup to work tomorrow and try there without any high power lines to interefere.

[BuriedEgg]

Thanks pinball! I'm going to buy a few of them tomorrow and put them in strategic places all over my wiring. Hopefully, that will kill off the noise hitting my wiring and I can get my game on!

[AndyWarne]

There is no anti-glitch delay built into the J-PAC deliberately, so that the response is maximised. The device sends a key the instant the contact is closed, which means it has no way to know if its a glitch or not. This I consider to be the best approach for this type of device.
Keyboards have a delay built in during which time the device can decide whether this is a glitch or not. I dont know if other suppliers interfaces have a delay.

In reality I have never until now heard of this being a problem, except with dance pads, which generate masses of static electricity. For the input to be triggered the amplitude of the EMI/RFI pulse would need to be huge so there must be some serious emissions from those power lines! The pulses are probably very short though, maybe too short to cause any effect on TV reception etc.

The workaround for this is to connect 100nF disk ceramic capacitors across each of the offending switches.

[newmanfamilyvlogs]

This thread has been a rather interesting read.... but if it's RF induction causing the key presses, why is it only those specific keys and not all of them? Are those wires just the longest / just the right length to 'tune' whatever is powerful enough to trigger them?

Another question, more for Andy.. Can you venture a guess as to how much power the "antenna" would have to be receiving in order to produce this effect?


And just some sheer curiosity for BuriedEgg, and a bit of a tangent: You might find it interesting to get a large loop of wire and build an 'antenna' that you connect to the microphone input of your soundcard and record a few minutes of 'air' at whatever the highest sampling rate your card can provide. You can actually monitor radio transmissions like this (though you need other equipment to monitor frequency ranges above what your soundcard samples at natively. What you'd be looking at is considered VLF or very low frequency). You might get lucky and pick up some harmonic or hint of what's there.
More info here: http://www.vlf.it/

[BobA]

The workaround for this is to connect 100nF disk ceramic capacitors across each of the offending switches.

Thanks Andy - Good to know, that is much cheaper then ferrite cores.

[BuriedEgg]

Well, I'm at work now. I've had the laptop test running for an hour straight now with no auto-typing at all, whatsoever. This means its the RFI/EMI in my house for sure.

I'm going to track down some of those 100nf capacitors mentioned here. I will post results.

I want everybody reading this thread to know that this is a problem with interference in my house, not a problem with the J-PAC itself. It says a lot about Andy's excellent support for his products that he is -still- actively troubleshooting this with me even though it clearly isn't technically his product's problem anymore.

I wish the company that built my home felt this way about customer support! ...or the company that built my car, or my HDTV, or my refridgerator, or my... oh you get the point. Way to go supporting the Home Arcade Electronics Community, Andy!

cotmm68030: I will definitely try that radio antenna experiment with my sound card and report back! That sounds really cool. I will record with Audacity and attach results.

[AndyWarne]

I used to have one that would occasionally go berserk and crash everything, and isolated the problem down to the JPAC.  (the line of BS I was fed at the time was I had clearly plugged it into my keyed connector backwards and fried something with the invisible power supply I didn't know was in my cabinet)

Well that would not have come from me. I have never once stated that any problem with a J-PAC has definitely been caused by plugging it in backwards. I might have suggested it as a possible cause amongst many others, in the interests of trying to assist with troubleshooting, but that would not have prevented us from repairing/replacing it which I hope we did in your case. I do not give my customers BS ever.

[BuriedEgg]

Okay, Preliminary Results Time and things look good!
(Photo attached at bottom in case anyone else runs into this problem in the future)

I found the Ceramic Disc Capacitors today but the store only had 4 of them. I bought them but decided to test with the Choke Cores first instead:


Attached 2 big Choke Cores to the main wires of the JAMMA harness as close as possible to edge connector. Was able to make only one full turn around each core.

Choke Core: http://www.radioshack.com/product/index.jsp?productId=2103979#



Then attached one smaller snap-on choke to the Kick Harness wires (the screw-connector wires on J-PAC) and one more to the extra disconnected wires (speaker/power/video) left over hanging loose from the main JAMMA harness:

Smaller Snap-On Choke: http://www.radioshack.com/product/index.jsp?productId=2103222



I ran the cabinet for about 3 hours last night and 2 more hours this morning with not one single phantom key-press or F2 reset.



I want to say this is resolved but I'll wait till Monday. It usually takes about 5 minutes to get the thing to start auto-typing keys but I've gone long stretches without it happening in the past.

Will update everyone on Monday.

The following pic is a side view of the 4 chokes on my wiring. The fourth choke, attached to the kick harness wires, is out of view right below the main edge connector.

I don't want to hear anything from anybody about the spaghetti mess in there! I keep a neat cabinet! Its messy because I've been tearing my nice neat harness up for the last week. Thanks guys!

[BuriedEgg]

Here is another pic revealing the lower portion where the screw connector wires are running thru the smaller choke. I was able to get an extra turn thru this choke.

(For anyone who may be experiencing this problem and wants to know how I applied the fix)

[BobA]

I am glad things are looking good for you.  Weird problems are the worst since logic does not always help get to the root of the problem.

[BuriedEgg]

I have new results to report:

After a almost the entire day of steady operation with the new Choke Cores attached to the wiring it finally did an auto-keystroke. It was P2-B6 that auto-triggered.

I wonder, if I go out and buy more chokes and just pile them on, if that would solve this thing. Is that ridiculous? Also, everybody, don't feel weird about making suggestions even if you feel it may seem random.

Just one auto-trigger all day long may seem pretty good. I know I may seem like a perfectionist. The reason I have to have the occurence of the auto-triggering buttons down to 0% is because just one single "F2" auto-trigger can ruin my entire progress in a game. That is a total show-stopper when friends are over.

Imagine having to explain that to a party of friends playing "Shadow Over Mystara" when the game resets in the middle of fighting Synn after collecting the "Sword Of Legend"! Yeah, that actually happened back in August. I was mortified...

[AndyWarne]

Eh, if you've already got all the wires RF choked, there's not much else to do there.

You could try the wrapping the JPAC in a towel and sticking in an anti-static bag, though.  

I don't think you'll ever completely eliminate the presses... that's a problem that's dogged me and others for years with those encoders.  I guess we'll all be shoving them in faraday cages ourselves (SO glad I was able to trouble shoot Andy's boards for him   )

I have never had one report before of phantom key presses with any of our boards. I have PMd you yesterday to ask for your email address so that I can remind myself what your problem was and how it was resolved. From your description it was an entirely different issue to random phantom keys. Also who are these "others" you mention?
If anyone else has ever experienced random phantom keys with our boards please post here.

Fact is, if other people were experiencing this issue I would have added a delay into the firmware which is a 5 minute task. Thats still an option as the firmware is downloadable but I dont want to introduce a delay into the button response on the basis of one user having an issue.

I have already suggested a fix for the extreme EMI/RFI problem that the OP is having but he has not yet implemented it.

[BuriedEgg]

Installed the Ceramic Disc Capacitors on "F2" and "P2-B6" switches.

No more auto-triggers!

That's it. Done.

Thanks everybody. I hope this helped anybody else living in an area with high RFI/EMI floating around. This was a really screwy problem that got me extremely frustrated but I'm glad in the end. I got to learn a lot from everybody here who contributed.

Thanks!

[Megaweapon]

Disregard

[ilpappa88]

Hi all,

I’m experiencing the same exact issues here but have a doubt on how to wire the 100nf capacitor. Would anyone be able to help me out?

Many thanks

[Zebidee]

Suggest you connect the 100nF (104) capacitors across the switches active and ground terminals. Probably easiest to solder them directly onto the microswitches if using standard cherry-type.

100nF (104) are typically available as ceramic, either flat disc (older) or monolithic, and they are non-polarised (orientation doesn't matter).