Weird issue with 13" K7000

[SirPeale]

So I'm bundling up my Arkanoid cocktail, putting the finishing touches on it.  I plug it in and leave it running for a bit.  When I check it, something WEIRD is going on.

The screen is all scrambled.  You can see the picture clearly, but it's like there's a magnetic ripple going thru it. At times it nearly clears, but...

I put my hand on the input wires and they were WARM.  I mean, *really* warm.  The connector itself was hot. 

I've moved everything away from the isolation transformer (I thought maybe I was getting crosstalk) and checked all my connections.  I just don't know why I'd be getting a hot input connector.   I'm fairly certain this is a monitor issue, not a board issue, but I won't rule it out.

[Ken Layton]

That's a 13k7800 which is of course part of the k7000 family. Sounds like your flyback is getting ready to conk out. Pull the chassis out and check for bad solder joints and recap it if it has not already been done.

[SirPeale]

Dying flyback would make the inputs hot?  The wires were warm all the way down to the board. 

I'm also not getting any other symptoms of bad FB - ie no arcing, no bad smell, etc.

Can this happen just like that?

Good thing I have a couple K7000 flybacks here.  I bought them for a song a couple years ago on eBay.

[Ken Layton]

Flyback and/or horizontal output transistor could be drawing more current than they should be.

How's the condition of the isolation transformer in this machine? Any electrical leakage to ground?

[SirPeale]

I recently rewired the whole cabinet because everything was connected to the iso.  It just wasn't safe.  So now the only thing attached to the iso is the monitor.

I'll have to meter the iso output; I attempted to do that before, and was getting a big fat ZERO on my meter.  Something's wrong with it.  Good thing I have like six meters!

[Ken Layton]

I would disconnect everything from the isolation transformer (power disconnected from cabinet of course). Now get a good sensitive analog meter and set it to the highest resistance scale like R x 10,000. Now measure the resistance from each primary terminal to transformer's metal frame (should show infinite resistance or open circuit). Now do same for the secondary side. Should also show infinite resistance or open circuit. Now measure from one primary terminal to one secondary terminal. Should show infinite resistance also.

If you get ANY kind of reading on any of the above tests, the transformer has electrical leakage and should be replaced. In other words it's not fully isolating anything.

Have you tried a different monitor in this cabinet to see if you have the same issues with a different monitor? If you do then it has to be cabinet wiring/connectors at fault.

[SirPeale]

Haven't tried other monitors yet, it's time consuming and I'm lucky I got five minutes to try and button this thing up as it is.  I guess I'll have to now, eh?

I have a cheap analog meter - somewhere - but it's a cheapie and I haven't used it in years.

I'm also not sure exactly what kind of iso this is.  On the output I've got several terminals, two of which the light for the coin door is connected to.  I'll reexamine it later.

[SirPeale]

*Finally* revisiting this issue.

I pulled the monitor from the cocktail and hooked it to my test rig.  I was expecting crap.  What I got was a perfect picture!

"Odd," I thought.  But it was time to try another monitor on the cocktail.  I have a 25" K7000 that was sitting next to it (needs a new tube, or rejuvenating, but otherwise works).  Plugged that in...also a perfect(ish) picture.

"Crap!" said I.  What changed when I rewired it?  I made it safer since everything was connected to the isolation transformer I had separate AC leads coming from the plug. 

I went upstairs and retrieved the monitor, hooked it up, but didn't mount it in the cab.  Again, perfect picture.  Now that's just weird.  I thought I'd just drop the monitor in place...

...and that's when it happened.  I touched a corner of the frame against the piece it mounts to by sheer accident.  SPARKS!  A lot of them.  And the picture freaked out like I mentioned in my OP.  I did it again just to confirm what I did, and it repeated.  I killed power and disconnected immediately.

The mounting frame of the cabinet is grounded to earth ground, which should be correct, right?  I've verified this with my meter's continuity meter.  The AC comes in the cab, and is split right away.  One lead going to the power supply, the other going to the isolation transformer.   Hot and neutral going to the appropriate tabs, and earth going to the frame of the transformer, which goes thru and connects to the metal frame of the cabinet. 

So I'm basically at a loss here; I can remove the earth ground, but I shouldn't have to do that, should I?

[Ken Layton]

And the monitor is the ONLY thing connected to the OUTPUT of the isolation transformer?

[MonMotha]

I'm not overly familiar with the power situation in Arkanoid cocktails, but is it possible that there is a fault in the DC supply somewhere?  Something might be "looking for earth ground" and finding it only via the monitor over the video lines.  That would certainly explain the hot connector and signal wiring.  Moving the DC power stuff off the iso output could have caused just this thing (as the iso could have been hiding this before).

[SirPeale]

I'm not overly familiar with the power situation in Arkanoid cocktails, but is it possible that there is a fault in the DC supply somewhere?  Something might be "looking for earth ground" and finding it only via the monitor over the video lines.  That would certainly explain the hot connector and signal wiring.  Moving the DC power stuff off the iso output could have caused just this thing (as the iso could have been hiding this before).

Anything is possible.  I'll have to trade out the power supply to be sure.

edit: I can just unplug it.  Duh.  When I rewired it I made sure to give it that possibility.

And the monitor is the ONLY thing connected to the OUTPUT of the isolation transformer?

Like I wrote above, this was originally not the case.  I made it that way.

[SirPeale]

Finally able to get to this today.  These days I have to wait for the neighbors to leave their apartment before I can work.  All the previous tenants in the last eight years were never around, so working downstairs was no big deal.  The new neighbors are there all the time, and with me puttering around outside their door it's like I'm doing it in the next room.  So out of respect, I wait until they're gone to work - which isn't often. 

Luckily today is a holiday, and they went to visit family for a few days.

Reinstalled the monitor, and it starts freaking out because it's connected to the metal frame.  I took a screwdriver and start backing out the screw that's connected to earth.  As soon as it was clear, the monitor had a crystal picture again.

So either something is wrong with the isolation transformer, or....?

[Ken Layton]

Or your wiring of the transformer.

How's your building wiring? Outlet properly grounded? Hot and neutral reversed maybe? Check your wall outlet for "leakage' voltage. Between the neutral hole and ground hole there should be zero volts AC (or as close to zero as possible).

Machine power cord in good shape?

Is it possible there's a problem with the "line filter" in your machine?

[Rickn]

I may have missed it, but what was the answer to ken's question "Is the monitor the only thing plugged into the Iso Transformer"

Rick N

[grantspain]

although i would expect it to blow the chassis fuse its not anything to do with the degauss coil rubbing against the frame
just a thought

[MonMotha]

To test my above theory, do a leakage test from your DC power ground to earth ground.  With the monitor completely disconnected, grab a large 10k resistor and jump DC power ground to earth ground.  Measure the voltage across the resistor.  Should not exceed ~25V (represents 2.5mA leakage).  Anything over 5V (500uA) could be cause for concern.  If that passes, then try again with a 1.5k to be sure.  Should not exceed 3-4V.  Measure AC and DC volts, though AC is of most interest.  Also measure the individual rails to earth ground, but should not be different other than the expected DC offset from power ground.

Another way to check this is to use a low rated fuse (25mA or thereabouts) in place of the resistor (no meter then needed).  If the fuse blows, then you have a problem.

This test can also be performed on the monitor isolated from the game electronics.  This verifies the integrity of the iso xfmr (also useful on monitors with SMPS).  Ken Layton earlier identified it as another suspect component for similar reasons.

You can also just try measuring DC power common to earth ground, but that can give misleading results.

If this test fails, then you have a faulty DC power supply which is shorted in some way.  In a linear supply, there is a short from primary to secondary.  In a SMPS, this usually would indicate a short in the transformer, but could also be a failed Y1 cap, shorted feedback optocoupler (very unlikely), or PCB flaw.  Could also be any other component that is connected to both DC and AC line, of course.

Other useful tests are to check the iso as Ken Layton mentioned (should read very high impedance from either primary to either secondary terminal) as well as the same check on the DC power supply transformer.

It is possible that the DC supply transformer (if this is a linear supply) is an autotransformer (primary and secondary are not isolated).  In this case, it was probably designed to run off the iso output.  When you removed it, that would cause exactly this problem.

[SirPeale]

Or your wiring of the transformer.

The wiring couldn't be simpler - AC in, AC out.  The transformer is an odd duck though... there are other lugs on it.  When I got it, the coin door lights were wired to  two of them. They're outputting ~6VAC.  I'll take a picture later.

How's your building wiring? Outlet properly grounded? Hot and neutral reversed maybe? Check your wall outlet for "leakage' voltage. Between the neutral hole and ground hole there should be zero volts AC (or as close to zero as possible).

Machine power cord in good shape?

Is it possible there's a problem with the "line filter" in your machine?

Building wiring is fine.  I don't have any problems with any other machine.  Nothing is reversed.  Power cord is fine, it's a computer cord.

Line filter might have a problem, but I doubt it.  Unfortunately it's built into the jack the wall cord plugs into.

although i would expect it to blow the chassis fuse its not anything to do with the degauss coil rubbing against the frame
just a thought

Nope.

I may have missed it, but what was the answer to ken's question "Is the monitor the only thing plugged into the Iso Transformer"

Couple replies above.  The answer is "yes."

To test my above theory, do a leakage test from your DC power ground to earth ground.  With the monitor completely disconnected, grab a large 10k resistor and jump DC power ground to earth ground.  Measure the voltage across the resistor.  Should not exceed ~25V (represents 2.5mA leakage).  Anything over 5V (500uA) could be cause for concern.  If that passes, then try again with a 1.5k to be sure.  Should not exceed 3-4V.  Measure AC and DC volts, though AC is of most interest.  Also measure the individual rails to earth ground, but should not be different other than the expected DC offset from power ground.

I'll experiment with some other power supplies to be sure.  I don't have any resistors handy to do those tests, so I'll have to try the fuse idea.

[MonMotha]

To test my above theory, do a leakage test from your DC power ground to earth ground.  With the monitor completely disconnected, grab a large 10k resistor and jump DC power ground to earth ground.  Measure the voltage across the resistor.  Should not exceed ~25V (represents 2.5mA leakage).  Anything over 5V (500uA) could be cause for concern.  If that passes, then try again with a 1.5k to be sure.  Should not exceed 3-4V.  Measure AC and DC volts, though AC is of most interest.  Also measure the individual rails to earth ground, but should not be different other than the expected DC offset from power ground.

I'll experiment with some other power supplies to be sure.  I don't have any resistors handy to do those tests, so I'll have to try the fuse idea.

That is kinda a last-ditch approach.  Before you do that, measure the primary to the secondaries of your DC power supply transformer.  If you've got anything but Hi-Z, you have either a shorted transformer or an autotransformer.  This assumes a linear supply.  If it's a SMPS, disregard that option, and you'll have to check leakage directly.  Check the same measurements on your monitor iso to be sure it's actually isolating.

[SirPeale]

It's a standard Peter Chou, so it's an SPMS, right?

I've got like 30 of them, so it's easier to swap it out than crack it open.

[Ken Layton]

This a Peter Chou with screw terminals?

Are the AC power hot/neutral wires connected to it reversed?

Does it currently have a jumper wire connected between Ground and FG?

[SirPeale]

This a Peter Chou with screw terminals?

Yes.

Are the AC power hot/neutral wires connected to it reversed?

I didn't realize it mattered.  I'll check that out.

Does it currently have a jumper wire connected between Ground and FG?

Nope.  Question about that though: why would you connect ground and FG together?  I have seen it done before, but it seems odd that you would combine an AC and a DC circuit together.

[SirPeale]

I puttered around with it today, and for the life of me I could not get it to freak out on me. 

I did find that the wires going to the power supply were reversed (hot/neutral) and swapped them, but that shouldn't have made any difference. 

The power supply doesn't indicate one way or another which terminal should be hot and which should be neutral. 

[Ken Layton]

So now everything is ok?

[SirPeale]

So now everything is ok?

So far it is...Hopefully it will never do it again.

[grantspain]

i hate problems like this as you are never 100% sure you have fixed it,its kinda like waving the magic screwdriver at a faulty machine

[SirPeale]

i hate problems like this as you are never 100% sure you have fixed it,its kinda like waving the magic screwdriver at a faulty machine

Tell me about it.  I'll keep running it (as much as I can, the neighbors are home now, dammit) but I'll bet it either doesn't do it again, or it'll do it as soon as  I've sold the machine.