Segai M21C XX hourglass screen

[trick72]

Hi,
I have a segai M21C XX CRT screen that has the "hourglass" look. The left and right sides or "curved" inwards
___
)  (
___

According to a manual that I have found (https://arcarc.xmission.com/PDF_Monitors/Happ%20Controls%20SE-M21C%20XX%2013%20or%2019.pdf), there is a "East/West Correction" pot on the board VR203. I have tried to turn the pot completely left or right or somewhere in the middle, but this has no effect at all on the actual image.
Could this pot meter be defective? Or is it somewhere else? Any idea what could be wrong?
Thanks

[buttersoft]

Your problem is commonly called pincushion distortion, or, as you've identified, East/West correction. It's normally corrected by an adjustment to the horizontal output circuit. However, that schematic you present has the components in the horizontal section starting with the number "3", so VR304, C302. It's the ones in the vertical section that start with "2", and you say you're playing with VR203?

I may just be raising this because i cannot even find VR203 on that schematic Can you point out where it is, please?

[lilshawn]

I'd look at the stuff around the LA7851... capacitors and resistors. it handles horizontal and vertical drive synchronization.

if you check out the datasheet for the LA7851, you can see what pins on the chip do what and focus your search for bad parts in these areas. (out of spec resistors or dried out capacitors.

[trick72]

Many thanks for your replies.
When you talked about pincussion I noticed there is a pot for that too VR304. With that I was able to correct the problem.

I have another problem with this screen I would like to address. There are at the sides of sharp lines some kind of "outline" lines. They usually have the same color as the original line, but with much higher contrast. It's like all graphics have some kind of "outline" in higher contrast to them. This is perhaps best explained with the help of some images I took.

When I lower the general brightness, the effect becomes much more prominent. Giving the screen more contrast and brightness lessens the problem a bit but not to a "normal" level.
Any idea what this problem is called? I was thinking about convergence, but not sure that it's the case. Also I noticed my tube (Segai) does not have any convergence rings at all.
Many thanks for your help.

[trick72]

I tried to hook up a real NeoGeo console with Metal Slug and there the picture is perfect. So it probably has nothing to do with the monitor, but with the PC. I tested on a mame PC running emudriver and groovymame.
I asked this question in the groovymame part of the forum.
Many thanks

[Zebidee]

"Ghosting" (what I call this) usually indicates an issue with signal transmission, essentially your cabling. This seems to be confirmed as when you use a console it works fine.

So... how do you connect your PC to this monitor?

[trick72]

"Ghosting" (what I call this) usually indicates an issue with signal transmission, essentially your cabling. This seems to be confirmed as when you use a console it works fine.

So... how do you connect your PC to this monitor?

It is an existing mame cab. It used to have a hantarex CRT which did not have this problem. Sadly that one broke and I have replaced it with this Segai. For the rest nothing changed on the cab or PC. It has a JPAC (ultimarc) connected from VGA to DVI on the ATI display card.
I just tried using a different vga to dvi cable but it's the same. I have a second working mame cabinet and I even tried to use the video card of that one into this cabinet but even with that video card (which works fine and does not have the ghosting on the other cab) it has the same ghosting on this one. So it's not a cable issue, or even video card issue. I even tried another JPAC and even that does not solve it. So it seems the Segai CRT has problems displaying this PC signal, while it works fine with a JAMMA/NEO GEO console signal. Very strange.
So I'm not sure now this is solvable.

[Zebidee]

Maybe it isn't the same "ghosting" I thought it was.

Maybe it has something to do with your video timings setup in CRT_emulator? Your Segai monitor might be very fussy.

I don't know what video modes your NEO-GEO uses, but you might look at trying different video modes in CRT-EMU (use ArcadeOSD) to see if it makes any difference.

Try using composite sync out of CRT-EMU.

Try bypassing the JPAC altogether for video. Probably need to rig up a breakout cable. Picture won't be as bright, but you'll get enough to see if the problem persists.

[trick72]

Maybe it isn't the same "ghosting" I thought it was.

Maybe it has something to do with your video timings setup in CRT_emulator? Your Segai monitor might be very fussy.

I don't know what video modes your NEO-GEO uses, but you might look at trying different video modes in CRT-EMU (use ArcadeOSD) to see if it makes any difference.

Try using composite sync out of CRT-EMU.

Try bypassing the JPAC altogether for video. Probably need to rig up a breakout cable. Picture won't be as bright, but you'll get enough to see if the problem persists.

I think what I experience with this CRT monitor is the "ringing artifact" which is covered for example in this thread:

http://forum.arcadecontrols.com/index.php/topic,162118.msg1713721.html#msg1713721

I see that you have posted a lot of responses there too. They are talking about putting a THS7314/7316 RGB amp between the pc and jamma signal to get rid of this issue.
So maybe I have to do this too? Strange that I did not have this problem with the hantarex monitor however and only seems to appear on the Segai...

[Zebidee]

Oh thanks, I'd forgotten about that.

The THS7314 chips do a few things that may have helped.

First, they have a sync-tip clamp or DC restoration circuit built into the inputs. This moderates any voltages above or below video spec, and "clamps" the signal to the flat/bottom part. This alone may have helped improve the signal.

Secondly, they include an (optional) LPF (low-pass filter) on each channel to cut out any frequencies below spec. He used that, and the filtering may have helped.

The LPF filter also introduces a small signal delay (~100ns or so). This is negligible in gaming terms. In real terms, it would shift the RGB image to the right by a fraction, maybe 0.5mm (depends on your CRT, how big it is).

In Bubblechump's case, this delay may have been exactly what he needed to bring the RGB into line with the sync. The large resistors (25k) he was putting on the RGB inputs to the THS7314 may have affected the filter/delay, allowing him to dial it in by trial and error.

The 500R he was putting on the outputs would have reduced the amped output signal from the THS7314, back down to something closer to what the JPAC expects.

This would fit neatly with my half-baked theory that your sync and RGB signals are slightly off somehow at the Segai end. The THS7314 and the LPF filter delays the RGB slightly, brings it back in line with the sync.

[trick72]

Thanks for your explanation. So i am going to order all parts I need. The amp chip is extremely small however. I never soldered that small so I hope it is going to work out fine...

[trick72]

So I managed to create the cable. The end result looks about the same like Bubblechump's although my resistors 25k are a lot larger so I could not fit them directly through the holes of the board.
But anyway, the image is much better now. Indeed with the 25K resistors as input and 500ohm as output the ringing is almost gone.
However, I notice some kind of "interference" now in the crt image. Some diagonal "transparant" lines going back and forth at various speeds. The image itself is stable but there are waving lines in them.
The VGA cable I cut and used was ofcourse fully shielded. But because I hacked it with the THS7314 and resistors this shielding is now much less "clean" and there is no shielding around my own wires and resistors.
I think this causes the interference on the crt monitor. Is there anything that could be done on this? Any tips on how I can improve shielding?

[Zebidee]

Well done!

Most likely cause of your waviness is signal interference, and quite possibly some part of your THS7314 amp mod.

So check all connections again.

Another possible cause of interference is the quality of your 5v power supply. If you were to look at any 5v supply on an oscilloscope you'll see some waves, some bumpiness. If too much , it can cause interference or upset the amp's operation and affect your image quality.

So, try an alternative 5v supply.

It will not hurt to add a electrolytic cap across the 5v and GND inputs, just like the 104 cap is in bubblechump's. I suggest try values 22uF to 100uF (I just read the THS7314 datasheet, it actually suggests exactly this!). Electro caps are polarised, so positive lead to 5v, negative to ground!

Failing all that, you could add a voltage regulator to the power supply, if so would suggest a LDO (low drop out). There are many out there, possibly already in your kit. The THS7314 needs at least 3v, there are many 3.3v LDOs available (for example, MCP1700-3302E, which is rated for up to 250mA output and available in a convenient TO-92 package).

You may also want to add a current limiting resistor to the 5v supply. I can't imagine the circuit uses more than 100mA / 0.5W so a 50R 1/2W resistor?

[trick72]

Good tips to get me going, Zebidee. Many thanks. I'm taking 5v from the PC power supply. (the power cables are yellow/black for 12V and red/black for 5V. So I spliced such a cable and uses the red/black for 5v. Maybe this PC power supply (I'm running mame on my cab) causes too much interference. I will try a seperate one and also the other stuff you suggested.
Thanks for your help

[trick72]

So it appeared the display problem I had was due to the video amplifier of the JPAC. It was too bright causing the ringing and other stuff. I hacked a VGA cable and used the R,G,B,V+Hsync and groud cables to directly connect to the CRT. Now the picture is fine and I don't even have to use the THS7314 anymore.
Only problem I am left with is the diagonal lines scrolling from bottom left to top right or visa versa in my picture. I had this also with the THS7314 between it, and I had it also with a normal VGA cable from the JPAC to the JAMMA. The image itself is stable and fine but there are lines "scrolling" in the picture. They very in speed. They usually start to scroll slow, go faster and slowdown again until they stand still, then start to scroll the opposite side. The lines are white and look like some ground issue but I don't think that's possible, as now it's just the PC video card directly to the CRT. I made a short video and uploaded it to this post (could not upload a MP4 video file so had to ZIP it), so you can see the lines. Mostly visible in the blue background of metal slug.
I have another mame cabinet with a hantarex screen and when I hook up this screen to the PC inside the "problem cab" I don't see the diagonal interference there.
So this would leave me thinking it is a problem with this Segai screen. However, when I hook up the original Neo Geo Console running Metal Slug, I don't see any interference at all on this Segai . The diagonal lines are not there.
The video I posted here is metal slug running on groovymame (crt emudriver) on a normal PC. I have the impression the video card of the PC is generating this disturbance lines. Because every time the resolution changes it seems these lines are "reset" somehow. They don't continue to scroll regardless of what is on the screen. They seem to reset when I change games or resolutions. Why they are not visible on the hantarex screen of the other cab, but are on this Segai screen is very puzzling to me, certainly because this Segai screen can display the NeoGeo console perfectly fine... I'm not sure if some kind of EMI filter or something would solve it. I tried to connect the PC power supply to a completely different broker circuit than the CRT screen power supply, but that did not help.
I hope someone can help on this. Thanks

[Zebidee]

This really does sound/look like a monitor grounding issue. Three ideas to explore.

1) Have a look how your Segai monitor frame is grounded.

Now this is going to be confusing (it is for me), but your logic or DC ground (this is your video signal ground) should not be connected to FG (full ground, or earth ground). FG is at your mains, DC ground is local to your device.

If your FG is also connected to your DC ground, then this can cause the interference you see.

Many people will say you should connect "exposed" stuff like coin doors, metal control panels, and the monitor frame to FG, and many manufacturers do this. Mostly for safety, in case of a short (although not really needed if the monitor frame is normally completely enclosed in a cab). However, monitor frame is usually also connected to the ground wiring behind the tube, which is then connected to the monitor's DC ground!

So, this is a long way of saying: if there is a wire connecting the monitor frame to FG, try removing it, see if it makes a difference.

2) Make sure your home-made VGA breakout cable has all the available grounds connected (pins 5,6,7,8,10).

In particular, make sure VGA pins 9 (5v), 11&12&15 (monitor ID pins) are not connected to your ground.

Make sure that the VGA shield ground is *not* connected to video grounds. This can cause interference. You may get connectivity between video ground and shield ground when cable is plugged into PC, but don't worry about that.

3) It may also be something not connected the right or expected way for the JAMMA. Are you still using the JPAC for connecting the controls?

Someone had a similar issue here recently, swapping game PCBs in a cab, but I can't find the thread. Anyway, I think they solved it by connecting relevant grounds on the JAMMA harness (or maybe the adapter loom, some games use non-standard harnesses).

[trick72]

Thanks for your reply Zebidee. I really appreciate it you take the time to help me out here.
There was indeed a ground wire connected from the monitor frame to FG. (There was a metal lip on the "plate" where the monitor system board resides to FG which had a ground wire on it). When I disconnected this the intensity of the lines is reduced by 75%. The scrolling and amount of lines is the same but they are much less visible now on some games but on the blue of metal slug from the video I attached earlier you can still see them. So you definately found the cause of the issue. Thanks for that! Now I assume there is still some path from FG to DC ground somewhere else... because the lines are still there being less visible. I see there is a ground strap (flat metal strap) around the monitor glass tube itself but not sure where this is going to... I think it is going to a ground pin on the monitor system board or neck board . But I didn't dare to disconnect it.

I double checked the pins of the VGA cable and they are as you described. The correct 5 pins are attached to ground, the others are not.
Lastly the JAMMA... I do use the JPAC for buttons. So the jamma wiring harness is still used. It is a version with a PS/2 keyboard cable going from the JPAC to the PC.  The grounds of the VGA cable are going straight into the 5 pin monitor cable (together with the R,G,B and sync), so I don't use the JPAC pin14 or JPAC ground pins for this video ground.

The PC power supply and cabinet (crt) power supply are on the same broker line, they share the same FG I guess. Could they interfere with one another?

[Zebidee]

There was indeed a ground wire connected from the monitor frame to FG. (There was a metal lip on the "plate" where the monitor system board resides to FG which had a ground wire on it). When I disconnected this the intensity of the lines is reduced by 75%. The scrolling and amount of lines is the same but they are much less visible now on some games but on the blue of metal slug from the video I attached earlier you can still see them. So you definately found the cause of the issue.

Part-way there at least

I see there is a ground strap (flat metal strap) around the monitor glass tube itself but not sure where this is going to... I think it is going to a ground pin on the monitor system board or neck board . But I didn't dare to disconnect it.

Yes, leave that grounding strap alone. The CRT can get stray voltages around the back, and the grounding strap siphons that off to DC ground. Leave it in place, and leave the wire connected to the neckboard.

The PC power supply and cabinet (crt) power supply are on the same broker line, they share the same FG I guess. Could they interfere with one another?

Maybe, but doesn't seem likely. Especially if you've already removed that FG wire from the monitor frame. Unless there is something else? The cabinet probably has an EMI/noise filter fitted near where the AC power comes in. These limit from noise/interference going in both directions (to and from the cab). If there isn't one, maybe it could use one. No harm done by having one, and they are cheap and easy to fit.

I confess I am running low on ideas.

Try running it with no JPAC disconnected? Just to see if it makes a difference. I'm still wondering if there is something about the way the JAMMA and JPAC are connected together.