WG K7000 series (now Wei-Ya 826HR) - replaced width adjustment cap, issues...

[Numbski]

I'm honestly not trying to make this difficult, thus why I did not turn the thing on.

When all of that is hooked up, the video ground is then tied to the frame on the picture tube.  You guys told me that was wrong, and I traced it to here.  I'll hook it back up that way, but it means that the frame of the picture tube is hooked to video ground.

[Numbski]

K.  Got it all hooked back up.  We have a nice, clear picture with completely hosed geometry.  What's next?

[Numbski]

If it's of any help, this is what it looks like without any video source, and the screen pot turned up slightly.  Notice how bright it gets bottom-right, and the pincushioning on the left.  Also, the artifacting lines top-to-bottom on the left remain without a video source.

[grantspain]

a bit scary some of this,problem is that you don't understand what we are saying mate.
video ground is only the signal input,the neck card ground is different(its an earth)-normally there is only one that connects to the dag but sometimes you have another that links to the frame,all depends on the monitor type

your next step is to post a picture of your screen with a grid pattern so we can see the geometry issues

[grantspain]

also i must add that although you may find this difficult and frustrating you are actually attempting to do things that many techs would not even consider.
we are all trying to help but you need to understand the basics first and more importantly safety aspects

[Numbski]

also i must add that although you may find this difficult and frustrating you are actually attempting to do things that many techs would not even consider.
we are all trying to help but you need to understand the basics first and more importantly safety aspects

I understand.  The two things on my mind when I went to bed last night were:

1.  Random internet people are both trying to help me fix my monitor AND keep me from killing myself.  I'm lousy at showing it, but I *am* thankful for what you guys have done here, which brings me to...

2.  Exactly how expensive shipping kegs of beer out are, because at this point that's what we're up to.

Anyway - I've re-posted the images below.  I've long thought about taking a course in electronics, and even bought an arduino board a month or two ago and started dabbling with that.  I swear until you mentioned it, I thought the terms earth and ground were used interchangeably, and that ground is ground is ground.  It all goes to the same place.  That battery comment finally put it together for me - and also brought back my studio recording days, trying to hunt down ground loops that cause speakers to buzz by having more than one path to ground.  Basically the arcade cabinet has 3 separate circuits going on.  There's the circuit that draws the AC mains, and has a transformer (the PSU) on it that creates our DC voltages.  There's another circuit that draws the AC mains, hits an isolation transformer and powers the monitor, and then the circuit(s) on dc power beyond the PSU.  I should have then realised that there were multiple ground lines right then and there - but again, for some reason my brain wasn't wrapping around this whole thing quite right.

From all indications, we have the multiple ground situation cleared up, and we have a clear image at least.

[grantspain]

now thats a good picture,i think the pincushion is being caused by the lack of width.
you must know by now that the width is adjusted using the width coil,what happens when you adjust that now?

[Numbski]

It moves in and out very slightly - not nearly enough to adjust it properly.

[Kevin Mullins]

Is that the only JAMMA board you have to test with?

Here is a Nokia Monitor Test program I use sometimes for testing monitors connected to my pc.
In other words....you should be able to connect your ArcadeVGA to the monitor and run this test from your PC to check the geometry and such as well. (instead of the jamma board)
Monitor Test

Check R43 which is right in line with the sync input, should be a 2.2K Ohm resistor.
Lower it's value and see if there is a change in geometry.
You can even short across it with a jumper to note any major change. (bypass the resistor)

[grantspain]

It moves in and out very slightly - not nearly enough to adjust it properly.

so you are back to the original fault which either a faulty width coil or a problem in the width circuit
unless of course this is a 20" chassis on a 25" tube

[Kevin Mullins]

unless of course this is a 20" chassis on a 25" tube

Good point ....

What is the P number on the main chassis?
Should look for something like a P447 or P538 or anything to that effect.

[Numbski]

I'm not at home, but I can say pretty emphatically that it's the right chassis.  I've had this monitor for almost 10 years - and it has worked properly the vast majority of that time.  Things only went awry when I went to install the cap kit.  The only thing that was wrong prior to that was that the picture was slightly too wide, in fact I have a picture of that too:



The image was slightly overscanned, and turning the width adjustment coil didn't appear to fix it, so I posted on these forums about a year ago, and Ken suggested I change out the width adjustment cap:

http://forum.arcadecontrols.com/index.php?topic=61421.0

So I ordered that cap, along with a full cap kit just for good measure.  That's when all of this began.

http://forum.arcadecontrols.com/index.php?topic=73028.0

Some time after this, I finally broke down, started calling around and found someone to repair the chassis, and turns out that I had shorted one of the transistors to the shield on the chassis, and that in turn had damaged some components, all of which he replaced for $70.  We got the wonky geometry, I fretted, ordered the Wei-Ya, took it back to the guy, who shipped it out to Vegas for repair, got it back and...ta da, we have the same picture again.  I'm not going crazy either - I got the same results hooking it to another tube he had laying around, so it's definitely not my tube.

Anyway...

My PC is running Linux - is that okay on the Nokia test?  I do have another board, a 6 slot Neo Geo, with only one cart at the moment, SvC Chaos.

[grantspain]

so you need to now checks all the cap values against the schematic in order to find if you have a wrong value cap in there

[Kevin Mullins]

My PC is running Linux - is that okay on the Nokia test?  I do have another board, a 6 slot Neo Geo, with only one cart at the moment, SvC Chaos.

Not sure about using Linux.... just try and run it on your computer without your K7000 hooked up, if it runs great, if it doesn't.... it doesn't.

You could try hooking the Neo Geo up just for reference that the image is still pincushioned.
Can't remember if it's the Neo Geo boards that have a slightly different video output voltage or something wonky to that effect. Was hoping you had some other generic JAMMA board laying around.

[grantspain]

snk require the vert and hor sync linked

[Numbski]

Actually - wait, I do.  I forgot, I have an Avengers board out in my garage that was installed in my Centipede cab when I bought it.   I'll grab that one and try it.

[Kevin Mullins]

Don't think that's standard JAMMA ... doesn't it require a Capcom to JAMMA adapter?

[Numbski]

I honestly don't know.  I just got home.  I thought it was CPS1 - thus Jamma, but it wouldn't be the first thing I was wrong about in this thread...

[Numbski]

P538

R43 does in fact measure 2.2k.  Jumping it now...

[Numbski]

Screen geometry retains the same distortion when R43 is jumped.

[Numbski]

NTest.exe will run under wine (just ran it on my mac), but I don't have wine loaded or configured on the arcade cab's PC yet.  If I had a BartPE disc laying around, I might be able to boot up and run it that way...

I think we can rule out the video source though - even when there's no signal, and there's just a white screen, the geometry is the same - the "whiteness" is even too narrow.

[grantspain]

i still reckon you have a wrong value component when you did the cap kit

[Kevin Mullins]

I think we can rule out the video source though - even when there's no signal, and there's just a white screen, the geometry is the same - the "whiteness" is even too narrow.

I have to agree with grantspain on double checking the cap values.
There's been known instances of a generic "K7000" cap kit having wrong values.
Especially when there are a few chassis that fall in that chassis family. P447, P538, etc.

I would check any cap labeled CAP LYT on this list against the ones you installed.
http://www.wgec.com/bom/P538.pdf

Double check that none are in backwards as well.
Might as well while you're looking at them.
(even though usually if they are they pop almost immediately)

[Numbski]

Will do - but probably not tonight or tomorrow night - it'll have to wait until this weekend.  Thanks for the pdf!