Main > Monitor/Video Forum
Wells Gardner D9200 - Horizontal Collapse?
Rocketeer2001:
Hey Shawn, I had read a bunch of your old posts about trouble-shooting monitors and remember you saying that a repeated clicking sound is the power supply restarting over and over. I just didn't know where to go from there.
I like the concept of narrowing down the problem by disabling parts of the board by cutting off its power, but as you said there is a risk involved. My board doesn't start clicking until at least 3-4 seconds later, so if I waiting that long with your method I'd be doing damage to other components. How would I know if the power supply wasn't clicking as a result of disabling the problem section, or if it wasn't clicking because the power supply section no longer worked? I suppose having a meter hooked up to read the voltages somewhere in the power section would solve that issue.
I should mention that it doesn't just click repeatedly, it also now makes a squeal or chirping noise along with the clicks.
Not sure I'm confident enough to start removing parts and turning on the power. I don't fully understand what all the parts do or their relationship to one another. Then again, this thing is destined for the trash if I try nothing, so what have I got to lose?
Tiger: I was having a hard time finding the samsung HOT part as well, so I went with the Fuji part FJL6920TU from Digikey. It's supposed to be a compatible replacement.
lilshawn:
typically if the power supply is clicking it's usually one of 3 things...
(in order of likelihood)
HOT's blown
SMPS rail caps old and dried out
vertical IC issue
most of the time your monitor will have been recapped at some point so generally the HOT is the issue, but if it hasn't been re-capped and it still has it's originals, chances are pretty good it's those. but honestly it could be either one.
the other issue that crops up is the vertical drive IC goes bad. you can usually tell if it's gone shorted if the resistors that feed power into it (usually 1 or 2 larger wattage resistors) are discolored or cooked looking. they will often test higher than what they are spec'd at due to the damage they have endured.
since you mentioned that the clicking doesn't start until a few seconds have elapsed, you are probably going to be looking at a system the fires up AFTER the computer boots and gets going. so something in the horizontal drive system or a pincushion compensation circuit...maybe something in the tube drive or more likely the fets switching the caps onto the horizontal (resolution frequency compensation capacitors) maybe even a sub power supply isn't working and it's dragging down the main SMPS.
try unplugging the connectors off the neckboard (leave the high voltage wires and grounds) and see if it sounds like it fires up. (4 or 5 seconds max) if it still goes to overload ticking, might be an issue with a shorted diode (like in the blanking system) or in the resolution frequency capacitor compensation switching system. (about 12 transistors/fets to check/swap/change/test etc)
<rant>
the D9200/D9400 is an obscenely over designed but poorly engineered chassis... it makes fixing them extremely difficult. the main problem being that it has so many independant and separately powered sub-systems. when one system fails, the others typically keep "running" causing more issues, which end up damaging parts in other systems. this makes fixing this chassis extremely difficult since you can damage new parts if you power on the chassis and the chassis isn't near 100% repaired. a still failed component can kill off a chassis in a matter of seconds... and you end up back at square one again.
last D9400 (not a 9200, i know... but near enough) I fixed required a list of 34 parts to repair, and about 8 hours worth of time over about 6 weeks to fix... and I kinda sort of know what i'm doing sometimes maybe but probably not over the last 16 some years i've been doing this... But even I have 3 or 4 D9X00 chassis in the chassis pile that i have given up on at one time or another.
with these monitors... my main advice is, if you are just a hobbyist electronics kinda guy/gal...and something completely obvious doesn't jump out at you, (like and obviously fried component or you change out the caps or a HOT and it still doesn't work) it is probably worth your time and money to send it off to someone for repair. Even if it takes 6 months to get back to you and costs you 300 bucks, it's still cheaper than the handfuls of parts and handfuls of hair and the weeks if not months of farting around you will lose trying the shotgun approach to only end up with a monitor that STILL doesn't work. Not being a negative nelly, just stating the reality of what these chassis can do to you.
</rant>
Rocketeer2001:
I replaced every single cap (about 78 of them) and I replaced the HOT twice. It could be blown again, but like you said it might test fine with a low voltage multimeter, yet fail when in the board with full power.
I think the limited trouble-shooting guide from Wells Gardner says this could also be a failed flyback. My understanding is that part is hard to come by these days.
I tested all the large resistors on the whole board and they all work now (I did replace one a while ago, I think).
I'll give that one method a try (unplugging the neck board low volt connections and turning it on), as that's a quick and easy test.
I don't interpret your advice as coming from a 'negative nelly' at all. I've been working on this thing every weekend (and some week days) for almost a year and it just seems to go from bad to worse. You sound like you know your stuff compared to me, and if it's still a struggle for you then my only chance of succeeding would be pure luck.
I tried to get local arcade repair shops to look at it and 5 of them said "NO". 1 place that is well known for being 'the' place to get your arcade fixed said they were too busy and even if they weren't they wouldn't take on this D9200 monitor because digital monitors are a real pain to work on. I tried emailing 3 or 4 online repair places and didn't hear back from any of them.
I'm going to try my 'Plan B', which involves using a TV as the monitor, coupled with a scan converter to downscale the EGA signal to a CGA signal. Might not look perfect, but it'll be better than nothing. And if that doesn't work, there is a guy about 5-6 hours drive away that has a rebuilt K7500 chassis with monitor that I could swap into the cabinet for $500.
Rocketeer2001:
Another day of disappointments, hurray!
I received in the mail the AVerkey iMicro scan converter. Lets me hook up the VGA 25kHz output from the graphics card on the game board and it'll output a 15kHz on S-video, composite, or SCART. I didn't have the SCART ready (the TV still needs to be RGB modded), so I tried S-video. Image is...fine...kinda muddy and has a slight jitter, but it was working...mostly. The image is shifted to the right a few inches, and cropped. I can center the image back into the screen, but I don't know how to 'uncrop' it.
On top of that, the light gun doesn't work. I pull the trigger, the screen flashes, but no shots are registered. I tried using the composite output instead, and same story. I'm wondering now if this converter is somehow messing up the signal the game board is expecting when you use the gun.
So close... :badmood:
lilshawn:
i posted a few years back about the ITs game boards and how the resolutions are not QUITE standard. they are close, but not close enough. not a problem for analog monitors where you can just stretch out or squash the picture coming in to compensate for the oddball resolution, but not close enough for most digital systems.
I had a similar issue with the GBS8200 trying to sync signals from golden tee machines back in the day, which is where the post came from.
i'm going to say the main issue with the gun not picking up is because of the scan converter... its introducing lag that is delaying the frames too much.
the way the system works is a little different from your average light gun game so i'll explain...
you pull the trigger, the computer sends a full white frame... at the beginning of the frame draw, it starts a timer... the monitor is drawing the screen in white, from top left across to the right...line by line...computer is waiting until the gun says it sees light. however long it takes till the gun says YES there is light now is whereabouts on the monitor it is pointed. then the scene is drawn again, once that's done, another full frame white is drawn and it counts again. this happens 10 or so times in a row (should the average be close enough together it could be as low as 5 or as high as 15) the counts are averaged together to get it about as accurate as it can get. that average number is where on the screen it's pointed and the computer acts accordingly.
problem with the scaler is that it introduces lag into the system. this gun signal timing until it knows for sure where your gun is pointed at is all done somewhere in the neighborhood of 160 to 450ms or anywhere from about 1/8th of a second to 1/2 second from the best case to worst case scenario.
so introducing any lag into a system where it's timing a portion of 16ms (how long a crt takes to redraw a single frame)... no matter how small... is going to screw it up. I think VERY likely the scan converter you have there is probably putting 100 or 150ms worth of lag into the frames. even the OSCC which is probably one of the best converters you can get that is often described as zero latency still has 50ms worth of lag...that is still just over 3 full frames... not going to work.