Gravitar Pcb... possible WG6100 Issues

LyonsArcade

Well-known member
Joined
Feb 9, 2006
Messages
19,367
Reaction score
3,854
Location
Rock Hill, South Carolina
Hi guys;
I've got a Gravitar here i'm trying to ressurect. I rebuilt the monitor tonight with new transistors/caps/etc. from Bob, and it has a pretty nice picture now.

The board is up and running, but the upper right corner of the game is messed up. At first I thought it was just some characters, but on the right and especially the upper right there are just ISSUES. I've taken a bunch of pictures. I checked all the Roms, they're fine. I checked the voltages on the board, they're fine. I swapped the two large Ram chips, and get the same issue no matter what position they're in.

Anybody have any ideas where to look next? I'm trying to decipher the schematics as we speak. Also I'm pretty sure it's a PCB issue, but does anybody know for sure that I've eliminated the WG6100?

When I put it in self test, and get to the screen that's supposed to show ram or rom errors, I DO get the low hum that it mentions in the manual, but the three rom chips it suspects there check out alright. The screen doesn't show any rom errors (that I can read!)

Any help is greatly appreciated! Board is extremely clean, never been worked on.
 

Attachments

  • 100_9078.jpg
    100_9078.jpg
    65.7 KB · Views: 82
  • 100_9079.JPG
    100_9079.JPG
    58.3 KB · Views: 64
  • 100_9077.JPG
    100_9077.JPG
    53.7 KB · Views: 60
  • 100_9076.JPG
    100_9076.JPG
    45 KB · Views: 56
  • 100_9074.JPG
    100_9074.JPG
    47 KB · Views: 54
Hi guys;
When I put it in self test, and get to the screen that's supposed to show ram or rom errors, I DO get the low hum that it mentions in the manual, but the three rom chips it suspects there check out alright.
Any help is greatly appreciated! Board is extremely clean, never been worked on.

What is the low hum supposed to indicate ? That one of those three roms is bad? How did you check them out? If they are socketed, it could be that the rom is ok but the socket itself is bad.
 
Yeah, it says basically when you get to the screen where it tests the sound and the switches, that if there's a low hum it may mean that one of the vector roms is bad. I'll have to dig through the sockets I guess tommorow to make sure they're o.k.
 
That looks very analog, so if you're sure it's the PCB, I'd check the output section (DACs and stuff). But are you sure the monitor is fine? I have a K6100 chassis with a similar problem (I haven't gotten around to fixing it). An oscilloscope would definitely help troubleshooting.

DogP
 
I'm not positive it's the pcb, but since the screens all have more or less severe issues I thought it looked that way. I just completely rebuilt the monitor, too but it could still be it I guess. You can fly your ship up into that area in the upper right and don't have as many issues with it's vectors.

I tried the pots, none of them adjusted it out.
 
Never trust your replacement RAMs. I just learned from a friend who fixed two Centi boards that I was running out of ideas on. I had replaced (and socketed) one of the RAMs and inserted one from my stock. Still error at that RAM, so I swapped in another one, still the same problem. I re-checked my socket work and it was 100% OK. I figured it must be a problem somewhere else on the board (buffers) and didn't work on it after that. Then my friend offered to have a look at them and one board was simply repaired by keep swapping in new RAMs. Turned out that 6 of the 15 RAMS I had bought from Arcadechips were bad !
Lesson learned: You can't trust old RAMs.

However, I've seen two vector RAM errors on my SW and the effect doesn't look like this. This looks like an analogue problem to me because of the bend vectors.
Maybe the 7815/7915 on the vector circuit ?
 
O.K., I'll check that too. As for the Rams, there's two as you know on the board, I would think if the vector ram was bad, when I swapped it for the other one on the board, it would cause all kinds of boot up problems, right? When I swapped the two, no change at all.

I'm going to check out that 7815 too then, thanks.
 
Hi,

If you had any bad rams please contact me and I will be happy to replace them. This can happen sometimes with the 2101's.

Thanks,
Mark Capps
 
Folks I finally got around to working on this again. My search-fu on the web turned up that if the BOTTOM right of the screen is distorted, ZD101 is often the problem. Well, on mine, the TOP right of the screen was distorted. All this info suggested to me that the problem might still be in the monitor, although I had rebuilt it.

Examining the schematics, I saw that zd101 was in the low voltage circuit that provides the -25 v for the monitor. I didn't have problems in the bottom right, so I wasn't looking for a problem here, only trying to decide why that would cause issues there.

That small circuit ties into the "-" line, and also the Y circuits.

There's a twin circuit next to it that ties into the "+" line, and also the X circuits. I thought that logically the problem may lie in that twin circuit, near the equivalent to ZD101. The 'twin' circuit had a ZD100, which tested fine, and a cap next to it. I changed the two caps (one for each circuit), and then noticed that Q100 tested bad in circuit. I had already changed this, I believe from Bob's kit. It's an MPSa06. Brand new, took it out of circuit, one side tested bad (instead of .4 to .6 voltage drop, I'm getting 1.3). Since it didn't short, it wasn't blowing a fuse or killing the +25; it was just making screwy stuff happen with the circuit it was attached to!

I didn't have an MPSa06 on hand, so I subbed in a 2N3904. Game is now up and playing with a full screen, all the dimensions are perfect, but the image is slightly shaky. I'm pretty sure this is because I'm using the wrong transistor in that low voltage section, but now I know. As soon as I get my hands on an MPSA06 I should be 100%.

Soooooo, if anybody has a similar problem, check that out. Also, don't always trust brand new parts!
 
Well I replaced the 2n3904 with the correct A06, but I still have a shaky image. I guess I'm going to look at the high voltage side of the monitor again. I'm getting 26.5v for the +27, i've heard that +28 is better, but I've rebuilt that low voltage section with all new resistors and transistors and diodes... the only thing I haven't changed is the big filter caps, could that be holding the low voltage down slightly? Or is it supposed to be +25 and +26.5 is fine?

Also would that voltage being low cause this little ripple I'm seeing in the monitor? Everything on the screen shakes slightly, all 4 quadrants. The B+ is set to 181.5v.
 
Well I replaced the 2n3904 with the correct A06, but I still have a shaky image. I guess I'm going to look at the high voltage side of the monitor again. I'm getting 26.5v for the +27, i've heard that +28 is better, but I've rebuilt that low voltage section with all new resistors and transistors and diodes... the only thing I haven't changed is the big filter caps, could that be holding the low voltage down slightly? Or is it supposed to be +25 and +26.5 is fine?

Also would that voltage being low cause this little ripple I'm seeing in the monitor? Everything on the screen shakes slightly, all 4 quadrants. The B+ is set to 181.5v.

The B+ should be as close to 180V as possible. If it is high, the HV output voltage will also be high. It's only 0.5% over nominal so it'll no big deal.

The metal cover for the HV box, when absent, has often been blamed to cause a shaky picture. On my 3 Atari vector machines, the difference caused by the cover is so negligible that I can hardly tell if the difference is imagined or real. What definitely can cause a severe shake (as in 1mm to 4mm of wobble) is the addition of an extra ground wire. Many vector machines of various manufacturers are adversely effected by the addition of a ground lead because it creates a ground loop. The effect was so bad on my Sega Star Trek that I thought that there was still something wrong with the game after I had completely rebuilt every section of it. Simply removing the extra ground from the monitor fixed it. If you want to make the ground connections super reliable, it's better to replace all of the terminals inside the molex connectors with new gold plated ones. Gold plated headers on the PCBs and gold trifurcon contacts in the harness housings are a major improvement. Such parts cost a bit more but they work much better and will last way longer.

Old capacitors will not bring the voltage down. They don't become additional loads, I mean not unless they short out. They will however lose some of their value (uF). The lower value leads to increased ripple on the DC supply. If the ripple is severe enough, it will cause visible effects on the picture. Since the deflection drive is a closed loop current amplifier, supply voltage ripple is automatically compensated for to a certain extent. There will always be a small percentage of ripple or supply noise that will get through regardless so, like I said, if the ripple is bad enough, it'll show up in the picture.

I recommend replacing the big caps (C100,C101) that were not included in Bob's kit. You'll find a very nice replacement part at Digikey as shown in my parts list here:
http://www.biltronix.com/arcade_electcomp_01.html#wg6100_rebuild_parts


Bill B.
 
Bill B.;
Thank you for your insight. Mine still has the metal cover, and I don't see any ground wires at all (I believe there's one in the 12pin molex connector to the deflection board, though). Also this wiggle is definately severe, and fast, it's not slow like a hum bar or a 'flag' wave or anything. Every vector is jittering...

I will get those caps, thank you. I did already replace c102,103 because I had those.. same with 800,801,802,803. Do you feel I need to do anything extra to the high voltage section? Or is this 'noise' more likely to be in the low voltage?

Thanks again for your help
 
I don't think the jitter is coming from the HV section. Yes, there is a common ground connection in the 12-pin molex connection.

As member Level42 previously mentioned, have you checked the 7818 (or 7815) and the 7918 (or 7915) regulator outputs on the game board in the video section? If either one of those is shorted through, the output will be unregulated (and too high) and that will certainly cause a lot of jitter.

Bill B.
 
No I haven't checked those, are those transistors? I gave a cursory glance at the schematics when he mentioned that before, but didn't actually see those, then went a different direction when I found the info about the common problem in the monitor messing up just one quadrant, which was similar to my problem. I'll check those 7815/7915's out, though just to be safe!
Thank you for your help.

Edit: I just checked over the schematics again and found them, in the schematics they're listed as part of the power supply (of course!). So I see them now, and yes they're transistors... I found them listed on your webpage too with a digikey reference, so I'm going to order a few to have around. Thanks for your help, I'll post back after I check all this stuff out and let you know what I found.
 
Last edited:
CAn you post a pic of the test screen now? Wiggle can be cause by incorrect resistors (the long white wire-wound ones). They should be non-inductive but are sometimes replaced with inductive.
 
You can't really tell anything by the test screen. The convergence is off slightly, but I'm not worried about that until I can get the thing to stand still for a while.

I ordered the filter caps for the deflection board, didn't change anything. I ordered some 7815's, they were out of 7915's... both tested good, and I replaced the 7815 anyways, still no change.

So so far what I've done:

Installed Bob Roberts kit, all bottlecap transistors, rebuilt low voltage section, etc.
New caps in low voltage section including filter caps.
Replaced 7815 on game pcb.
Replaced Big Blue
Cleaned edge connector on game pcb, verified and reseated all roms.
Cleaned fuse holders & fuses.
Checked ground on power cable
No extra wires going to the chassis.
Made sure cage was on hv unit


I can order those resistors, but this is a game that I'm not so sure has ever been worked on. It's in near mint condition, and spent the last 15 years in a family's gameroom. If they would have replaced those resistors wrong, surely they would have immediately had image issues and took them back out, I can't imagine they put it out on the street with the image constantly shaking. It's slight enough that you can't picture it, but it's constant, almost like when an older television wasn't tuned in quite right, the whole image has static running through it moving everything.
 

Attachments

  • 100_9873.jpg
    100_9873.jpg
    91.7 KB · Views: 15
  • 100_9874.jpg
    100_9874.jpg
    96.9 KB · Views: 9
Alright, so my latest is this. I ordered the two resistors that are supposed to be non inductive for the deflection board, and replaced them, no change.

Then I ordered a rebuild kit for the ARII pcb, rebuilt it, no change. Swapped in another one I had in a millipede, too, no change.

So then I replaced all the pots on the Gravitar pcb, and also the two filter caps that the 15v and -15 run through.

No change.

Also, from before:

Installed Bob Roberts kit, all bottlecap transistors, rebuilt low voltage section, etc.
New caps in low voltage section including filter caps.
Replaced 7815 on game pcb.
Replaced Big Blue
Cleaned edge connector on game pcb, verified and reseated all roms.
Cleaned fuse holders & fuses.
Checked ground on power cable
No extra wires going to the chassis.
Made sure cage was on hv unit


Anybody got any more ideas?
 
Just for the record, I finally fixed this.

I ordered another capkit, and also Bob Robert's 'extra parts' kit.

I focused on just the high voltage since lots of people were saying that's what it sounded like. Mark Spaeth and the guy at Arcadeshop both 'opined' that it didn't sound like a low voltage problem, and then advised me not to buy their low voltage boards. Both straight up guys if you ask me. Yeah, you read right, Mark said not to buy his board because it probably wouldn't fix it, then when the guy at Arcadeshop mentioned his, and I told him my problem, he did the same thing.

So I took the cage out, started replacing all the things I had already replaced; also I put the higher value resistors in that the extra kit had.... this was about the time I noticed lonely 'ol C922. C922 apparently is only on some chassis; so the capkit doesn't have that cap. It's just a 10uf 63 volt. I had tons of extra caps around (don't know why I didn't switch it before)...

So I put it back in after all that, and the picture is rock solid. So It could have been one of the transistors, caps, or resistors I replaced for the second time, but my hunch is that it was just C922 hadn't been changed the whole time.

So, for posterity: CHANGE C922 EVEN IF IT ISN'T IN THE CAPKIT, IT'LL SCREW UP THE MONITOR!!!!
 
Last edited:
Back
Top Bottom