Betson Tri-Sync Monitor Help

[XtraSmiley]

Hi guys,

I'm running a Betson Tri-Synch digital montior in my MAME set up and it's 5 years old, but only had a few hours of use on it. This problem was there from day one though.

The far right side of the screen (by the bezel) is squished looking. It's really noticable on higher resolution games and of course anything that scrolls from right to left looks terrible, like TMNT, Simpsons, etc. In fighting games, the second player looks really skinny when all the way to the right.

I've tried all the monitor adjustments, including the secret menu (where you hold the OSD button before turning it on) and the problem can't be fixed.

It's not super bad, like very squished, but noticable enough that it bugs me and shouldn't have shipped from the factory this way. I wish I had just gotten a WG9200 like everyone else!

Anyway, any ideas? I know "cap kit" seems to be the one stop shop answer for this, but the caps should be brand new! Also, I would have to have someone do it for me and if I can maybe just find the one cap that is causing the problem maybe I could do it.

Thanks for any help.

EDIT:

I see from here:

http://forum.arcadecontrols.com/index.php?topic=80967.0

and here:

http://forum.arcadecontrols.com/index.php?topic=26178.0

That it seems to be common problem, but no one ever posted how they fixed it, I hope they did...

Mine seems to be worse than everyone elses!

Last but not least, can someone one with the stock operator's setting please post their original settings, I think I've f'd mine up! Thanks.

[MonMotha]

This is a known problem with no real fix.  It's generally only present on 15k modes.  You can get around it by upscaling (e.g. line doubling) to 31k, but that has other implications.

There is a horizontal linearity control in the service menu, however it will affect the whole screen, not just the edge.  You may be able to arrive at a better compromise than you have now, though.

[XtraSmiley]

This is a known problem with no real fix.  It's generally only present on 15k modes.  You can get around it by upscaling (e.g. line doubling) to 31k, but that has other implications.

There is a horizontal linearity control in the service menu, however it will affect the whole screen, not just the edge.  You may be able to arrive at a better compromise than you have now, though.

Thanks, but why? I mean, what is causing the problem, surely it can be adjusted a bit right? Is it a yoke issue? PCB?

[MonMotha]

It's a design defect with the monitor.  It's much worse on the later "Dynaflat" (-DF) model with the perfectly flat CRT.  The earlier model ones with the "flat" (-F model) CRT don't do it as badly.

They do ship poorly adjusted, and there are some things you can do.   In one of the threads you linked, I described a trick using pincushion, pinbalance, and trapezoid to help with the squish a bit at the expense of straight edges.  Again, you can also play with the horizontal linearity control, but as promising as it sounds, it won't help as much as you'll want (you can make that side of the screen somewhat better, but it'll start to squish the other parts of the screen in weird ways).

Basically, the horizontal deflection waveform isn't what's needed to get a linear sweep across the tube face.  Short of mucking with the S-correction caps (which can really futzor things if you're not careful), I'm not sure what to go playing with.  It's not something in need of repair - they just are this way straight from the factory.

[lilshawn]

the superflat monitors need all sorts of adjustments on the fly during the scanning of the line. it needs to slow down...speed up, then slow down again due to the extreme angle which the beam has to turn before reaching the screen front. I'ts pretty gnarly if you look dowb at the tube from above and imagine a line going down the neck and doing a hairpin turn at the yoke to hit the extreme left and right sides.

like what's been mentioned, there is some correction caps that adjust the timing of the speeding up and slowing down of the scan rate but you can really mess things up in a hurry and it doesn't take much.

all this is due to the "universal" nature of the chassis they use. It's basically a fits-all for everything. I'm sure that chassis looks perfect on a normal "round" tube. looks okay on a "faux flat", and looks terrible on a flat tube.

[XtraSmiley]

That blows, but I should say that mine is not the flat type, it is curved. Dang, this really blows.


Ok, can someone still post the default service settings?

[grantspain]

you can fix that issue by changing a high voltage poly cap in the horizontal circuit
if you do this then 31khz will display issues

there is a tutorial on arcadeotaku forum where a member has done the modification and put in a switch so that both resolutions display correct geometry

[MonMotha]

The monitor should already be switching caps in and out between 15 and 31k mode.  There may be a way to make it automatically switch the replacement caps (or even take advantage of the existing PCB features and just put the replacement in a different place).  Happen to have a link to this fix?  I never bothered because I was always afraid to go tweaking those caps too much for fear of breaking other stuff.

[XtraSmiley]

you can fix that issue by changing a high voltage poly cap in the horizontal circuit
if you do this then 31khz will display issues

there is a tutorial on arcadeotaku forum where a member has done the modification and put in a switch so that both resolutions display correct geometry

Can you link this? I'm not sure I understand, will replacing the cap fix the issue? If its that easy why doesn't everyone do it?

[MonMotha]

The challenge is knowing what value to replace it with.  If you get it wrong, you can pretty well destroy the horizontal output section (though probably just the HOT itself).  Apparently somebody has worked out the proper value, which I would also like to know.

[grantspain]

http://wiki.arcadeotaku.com/w/Rodotron_666#Side_Compression_in_15KHz

[XtraSmiley]

http://wiki.arcadeotaku.com/w/Rodotron_666#Side_Compression_in_15KHz

This doesn't help me, I have no idea how to do this. Also, it's not for my Beston.

Does this info mean I could take this monitor to a repair guy/shop and they would be able to fix the side compression issue?

[iamnaeth]

I shipped my control board (i think that's the right term) to Betson to work on this.  It didn't come back any better.

[XtraSmiley]

Well, being compressed on the right is bad, but I can deal with it. It's the damn scrolling games that are very hard to deal with. It's really bad on these!

Anyone in the VA/MD/DC area that can take a hack at fixing my monitor?

[MonMotha]

Certainly Betson won't "fix" this.  It's not "broken" as far as they are concerned.  They *all* do this, straight from the factory.

I can look up the equivalent cap in a couple days (traveling right now and don't have time).  On the KT-2914 chassis, there are already switched caps per mode, but it's possibly they just flubbed up the value on the 15k one.  Would be nice if a simple component swap would fix it.

Fair warning: there's a chance you'll break something (semi-permanently) with this mod.  Apparently it works well on the Wei-Ya, but short of somebody testing it on the Kortek, I'm certainly not qualified to tell you if a particular value will work right or not.  CRT deflection electronics are somewhat of a voodoo art to start with, and I don't exactly have any design notes on this thing (just a schematic).  One of the CRT old-timers might be able to give you a bit more of a hint.

[8BitMonk]

Is this only a 'known issue' on a small number of defective units? I have a Betson Imperial Tri-sync and have never noticed this problem or had any issues with it. 

[MonMotha]

It seems to affect pretty much all of them, from what I can tell.  The DF (totally flat tube) is way worse than the F (mostly flat) model.  It only occurs when running at 15kHz scanrates.  24-38k are fine, so if you're upscaling everything in an emulator, you'll never notice.

The fix discussed for the Wei-Ya/Rodotron makes sense.  It just can't necessarily be applied directly to the Kortek without some possible consideration.

[8BitMonk]

Maybe I'm missing something, I run all of my games at 15khz in their native resolutions using an AVGA card with no scaling and haven't noticed any issues.

[MonMotha]

Then you, sir, are apparently lucky.

Makes me curious what they changed along the way.  Every one I've ever seen running at 15k seems to exhibit the problem to some degree, though it's somewhat variable.

[8BitMonk]

XtraSmiley did you physically calibrate your monitor using a monitor test tool like the Nokia Monitor Test? I seem to remember having some display issues before I did this. I also use AVRes with mame to automatically set custom .ini files with the most appropriate res for each game. Trying to think of things I've done that you may not have.

[XtraSmiley]

Thanks, I don't know what the test tool is, but I have the individual ini files thanks to the utilities off of the Ultimarc page.

Many games that scroll just look terrible. I'll look into this tool. Do you have a link?

[iamnaeth]

Your games that scroll, does it look like things are kind moving over a waterfall or cliff on once side of the screen?

[XtraSmiley]

Your games that scroll, does it look like things are kind moving over a waterfall or cliff on once side of the screen?

Sort of. It's like on the right side, an item that should be 2 inches wide is only 1 inch (compression) and as it scrolls to the left, it grows in size until it's the full 2 inches. Starting out compressed looks bad, but I can deal with that (like in games with stationary screens, example Tetris). It's games where you move from left to right (TMNT, Double Dragon, Golden Axe, etc) where the objects grow to the correct size, looks weird and is very distracting.

[8BitMonk]

Hey, sorry for the late reply, I forgot about this thread. You should definitely use a monitor test tool to calibrate, absolutely necessary. I wouldn't be surprised if this is your issue.

I don't have a link for the Nokia test tool, just google 'Nokia Monitor test'. It isn't Win 7 compatible, but there are others that are http://www.prad.de/en/monitore/testsoftware/eizo-monitor-test.html and http://www.passmark.com/products/monitortest.htm. They just basically show patterns and colors so you can adjust positioning/brightness/contrast using the physical controls of the monitor. Use the little keypad connected to the wires on the back of your monitor to make adjustments.

As for the scrolling issues you're speaking of, you may just be getting some screen tearing because the refresh rates aren't matched perfectly. You can usually get rid of this in games by turning on/off nowaitvsync and/or tripplebuffer in the custom ini file for the game.   

[MonMotha]

The equivalent cap on the KT-2914 appears to be C709.  The schematics for both models are terribly drawn and really hard to follow, but C709 is definitely an S-correction cap on the KT-2914.

Stock value indicated on the KT-2914 schematic for C709 is 0.56uF.  Note that, at 15kHz, C705 (0.22uF), C702 (0.15uF), and C710 (0.27uF) will be paralleled with it.  C709 is switched out at scanrates higher than 15kHz, so changing the value of this cap should only affect 15kHz (i.e. no manual switch needed).

I just checked my KT-2914F (note: NOT the DF), and it is indeed a 0.56uF.  If you have a DF, and this value is the same, then they probably didn't even bother to check/mess with it when they changed the tube type.

Now, as to what value you should change it to, I have no idea.  I guess you can try bumping it up and seeing what happens if you like to live dangerously.  These caps not only affect the shape of the deflection waveform but also tune the circuit, so going too far off will break things (potentially somewhat catastrophically).