What’s wrong with my V2000/G05 B&W vector monitor?

I don't know that I remember ever seeing three of the metal clips on a frame before, usually its just the outer two. But maybe I just never noticed the center clip.

I just checked the frame, its an Electrohome G05-802 frame from Jan '81
 
Also, I'd try the suspect tube in a different monitor (if you have one), before being certain the tube is the issue.

This is one of those cases where just because you see a problem go away, you can't be 100% sure the problem is what you think it is. It's easy to get fooled. (This applies to cases where you think a chip is bad, too.) It could just be a bad connection or the fact that you were in there touching things that made the problem go away.

So you not only want to make the problem go away, you also want to make it reappear in a totally different setup if possible, using the suspect part (as long as you know doing so won't damage anything, which in this case it won't.) That way you know 100% that the part you think is bad, is bad.

This is one of those 'good practice' things you learn if you do enough repairs. Especially for tricky troubleshooting cases (or for expensive parts like tubes, Pokey chips, etc), it pays to do the extra work to sanity check your conclusions when you think you've figured something out.
I cleaned the crt pins and hit the socket with deoxit but it didn't help. The first time the socket had ever been removed from the tube was by me very recently. I'll hang onto it to test in another game later if I pick one up but will be searching for a new tube 🙃

Also the screwdriver trick didn't work and I don't have a cr6 adapter for my rejuvenator.
 
Last edited:
I don't know that I remember ever seeing three of the metal clips on a frame before, usually its just the outer two. But maybe I just never noticed the center clip.

I just checked the frame, its an Electrohome G05-802 frame from Jan '81

I'm pretty sure all 802/805 and v2000 frames have three clips. I've just never run into a problem with them shorting anything out, and as a result I've never looked closely to see what each of the clips comes into contact with.

It's an interesting question though, as you'd think it would cause more of a problem. I've *definitely* had the other issue cause a problem (where the board rubs against the back cross-bracket of the frame, next to the upper screw.) If you don't install that screw, and have it in an AD where gravity pulls the board down and into contact with the frame, the rear cross brace will short one of the traces out. But only if that particular board has its trace right at the edge of the board, and not all do. But I kept wondering why I was sometimes seeing sparks on powerup for some boards, but not others. :)

But I've never looked at the clips, mainly because it isn't easy to see underneath the board.

I just grabbed a spare 802 board from the shelf, and noticed it's a later rev than yours. As a related side note, people misread these all the time, because there are actually four version (aka 'issue') numbers corresponding to four separate aspects of the board. Each aspect has its own issue number, and they are all independent.

There is:
- A 'Foil Pattern' issue number on the solder side (for the actual copper trace layout)
- A 'Solder Mask' issue number on the solder side (for the pattern/shape of the mask)
- A 'Comp Symbol' issue number on the solder side (for the component number silkscreen)
- A 'Comp Layout' issue number on the parts side (for the part symbol/number silkscreen on that side)

See pic below for the first three, outlined in red. The other one is just printed on the parts side, similar to the Comp Symbol number.

You can get different combinations of all of these issue numbers. I actually record all four numbers on every board I refurb, and have a bunch of data. But I've never analyzed it to see how many unique combinations there are, but it's more than a few. The main point is, if the parts side says 'Issue 5 Comp Layout', some people see that and call it 'Issue 5', but that's not the whole picture. You have to look at the other numbers too.

The reason I mention this is because for the board I grabbed, it's a Foil Pattern Issue 7 (versus your Issue 6), and it looks like they changed the layout around that D600, to make more space (presumably to increase clearance around that clip). See the blue oval in pic below.

That's super interesting, as I often wonder why they did so many versions of these, but have never figured out all of the differences, let alone what motivated them. But this one seems to make sense, given what you've pointed out.

ISS 7 foil.JPG
 
Thanks to Andrewb for selling me a working asteroids pcb- that was clearly the issue and his pcb got the game working right away-
 

Attachments

  • IMG_1172.jpeg
    IMG_1172.jpeg
    1.5 MB · Views: 7
  • IMG_1173.jpeg
    IMG_1173.jpeg
    1.1 MB · Views: 7
Is there any way to verify flyback operation before it hits the HV diode? Can I stick an HV probe in the boot and get a voltage?
 
Is there any way to verify flyback operation before it hits the HV diode? Can I stick an HV probe in the boot and get a voltage?


Technically, yes. Practically, no.

The HV before the diode is HV *AC*, not DC. And to probe it you'd need to have the boot off and cup exposed (assuming this is an 802 or v2000), which is very dangerous, as it's going to want to arc to anything. You won't be able to just stick the probe in the boot without holding the boot, which isn't possible. So it's not really practically possible.

Better to just assemble the whole thing, diode and all, put it in the tube, and test at the anode. Usually when the diode is bad, you'll get *some* HV (like under 2kV). And/or the diode will get very hot within a few minutes.

If there's absolutely zero HV, it's possible the cage isn't running at all due to some other bad part (transistor, resistor), The flyback is the least likely thing to go bad. Not impossible, but it's rare. And the only way to be 100% sure is to swap it into a known-good cage.

This is a case where it helps to have another known-good cage to borrow good parts from, to narrow things down when just popping a new diode in there doesn't fix it. (Though it's good to always have a spare good diode on hand at least, if you don't have another good 802 or v2000 cage to borrow parts from.)

What problem are you having, and what have you tried/replaced so far?
 
It was a working V2000 that I painfully watch fade to black while I was playing it. All the electrolytic caps have been replaced along with the bottlecaps. I can hear deflection chatter, but I get no HV voltage at all. I cleaned the cups with acetone and vinegar soaked and sanded the springs for the diode. I soldered the ends of the diode to the springs, but I think upon reassembling it, the diode pin broke off, because it was broken when I pulled it back out. I've tested Q903 and Q900. I've ordered a pair of H1812 from ArcadeShop, which are actually #2 and #3, because I broke the pin off the last one I ordered. These things are sensitive.
 
It was a working V2000 that I painfully watch fade to black while I was playing it. All the electrolytic caps have been replaced along with the bottlecaps. I can hear deflection chatter, but I get no HV voltage at all. I cleaned the cups with acetone and soaked and sanded the springs for the diode. I soldered the ends of the diode to the springs, but I think upon reassembling it, the diode pin broke off, because it was broken when I pulled it back out. I've tested Q903 and Q900. I've ordered a pair of H1812 from ArcadeShop, which are actually #2 and #3, because I broke the pin off the last one I ordered. These things are sensitive.


If it's fading out, you may be losing power to the HV cage. But that you could measure at the input to the cage, fairly safely.

For all of these b/w monitors, one thing that can happen is bad connections on the headers. Take a flashlight and shine it into all of the header connector plugs. Make sure the contacts are all lined up with each other, and that none are bent or squished in. When they are, they will make intermittent contact with the pin, and open up when they warm up.

If you're saying you watched the monitor go dark and it's permanently that way, something else may have failed in the cage. You can replace the two small TO-92 transistors in the cage pretty easily (they are both MPSA06's, I think). And test the cage-mounted ones, which you can do by just sliding off the connectors. (I always pull both of these and hit the legs with the Dremel wire wheel brush, as they are often pretty oxidized. Then put some DeOxit on them before putting the connectors back on.)

Also, DeOxit the pot and work it back and forth, as the pots in v2000's can be janky sometimes.

Also, do you have neck glow? Just to doublecheck. Fading can sometimes mean a bad heater connection (though you should still have HV in those cases.) Fading can also be caused by bad transistors in the Z circuit on the deflection board (which I've covered in other posts), but again that wouldn't cause no HV.
 
Pulled the T0-92 transistors and tested them. Both seem good, ~.640 with + on base with - on collector and emitter separately.
 
Pulled the T0-92 transistors and tested them. Both seem good, ~.640 with + on base with - on collector and emitter separately.


In my experience these small-signal transistors can occasionally go bad, but test ok with a 2-diode test. They can just get weak in-circuit.

They're ten cents each, so no harm in replacing them in this case, if you aren't sure, and don't have a curve tracer. That's what I'd do next in cases like this. Then swap the two cage-mounted transistors with ones from a known-good cage. Then check all caps (not just electrolytics) with my ESR meter and compare readings to a good cage, side by side. Then swap the pot. Then swap the flyback with a good cage, as a last resort.

You can also try toning out the coils of the flyback (both for the coil resistances, and making sure coils aren't shorted), per the schematics. But that also isn't a silver bullet. But the couple of flybacks I have seen bad have read slightly different from a good cage.

Swapping the flybacks with another good cage is a pain (you need a hot iron and good desoldering gun, and you have to desolder then resolder two flybacks to properly swap them). But it's the only way to be 100% sure a flyback is bad, once you've ruled everything else out.
 
Update. I eventually found that I had wired Q900 and Q903 backwards. Now I'm getting a faint image that balloons to nothing. Glad my flyback isn't dead. I assume that although the HV diode is new, the soldering I did to the springs is not sufficient? Going to resolder it now.
 
Update. I eventually found that I had wired Q900 and Q903 backwards. Now I'm getting a faint image that balloons to nothing. Glad my flyback isn't dead. I assume that although the HV diode is new, the soldering I did to the springs is not sufficient? Going to resolder it now.

Correct.

The HV diode is probably also getting hot, which is another sign that the connections are not solid.

One trick to soldering the springs to the diode is to get a small jeweler's file (or some 1000+ grit sandpaper), and file the ends of the springs to shiny bare metal. (And do the same to the diode leads). That will get the solder to 'wet' to the steel springs more easily.

If the solder isn't actually wetting to the steel, it isn't making solid electrical contact. And in that case the joint has some resistance, which basically turns into a heater when you pass15kV through it (even at a small current).

Also make sure the metal cups inside the boots are perfectly clean. I stick Goof Off-soaked Qtips in them until they come out clean. Then take a flashlight and look down inside them, to verify there isn't any debris left inside them.
 
Should the metal cups be soldered to the wire? I used a nail file on the springs, but I think it still isn't perfect.
 
Should the metal cups be soldered to the wire? I used a nail file on the springs, but I think it still isn't perfect.


The cups are actually spot welded to the anode wires. Then the springs sit inside them by friction, no solder involved.

There's no reason to remove the cups, unless you tried to replace the boots (because the cups need to be removed in order to slide the boots off, and that requires cutting the wires behind the cups.) However you normally don't want or need to take the boots or cups off.

If you did cut the cups off (or if the wires were broken behind the cups), then the wires need to be soldered back to the cups.

This is tricky, as it requires filing out the insides of the cups (in order to get the solder to stick). I use a small diamond-tipped Dremel grinding bit. Then you have to cut the wire and file the tip of it bare. The cups have small holes drilled in them that you can stick the wire through, and then solder from the inside. And you want to use as little solder as possible, as you want the inside of the cup to still be 'cupped', so the spring still seats in it. And you have to clean all flux off after, because it needs to be spotless.

So yeah, you don't want to remove the cups (or the boots) from the wires, if possible. Doing so makes an already tricky repair twice as tricky.
 
Le sigh. I soldered the flyback cup because it was barely hanging onto the wire. The spot weld definitely wasn't there if it ever was. More like hole and wire hooked through the hole.
 
Le sigh. I soldered the flyback cup because it was barely hanging onto the wire. The spot weld definitely wasn't there if it ever was. More like hole and wire hooked through the hole.


You mean the HV diode cup on the flyback side of the diode?

The wire is normally passed through that hole in the up, and then spot welded on the inside. If you look inside the other cup, you should see what I mean, as it's done the same way.

Maybe the weld came loose in your case. But yes, that joint needs to be perfectly done as well. I actually use 2% silver solder on those when I do them, as it flows and sticks a little better. That isn't necessary, but what is critical is that the joint be perfectly clean, and solid.

One key is to file the stranded wire bare and shiny beforehand, as the copper is usually oxidized, even though it was inside the insulation. That layer of oxide will greatly interfere with the solder's ability to wet to it. So filing at least the outside of the strands will make the solder stick much better. I just file it against my finger, then re-twist the strands. It doesn't have to be perfect, but having some of the strands be bare and shiny is better than none of them.
 
HV cup on the flyback side. Where can I source the 2% silver solder? Amazon is a confusing mess.
 
HV cup on the flyback side. Where can I source the 2% silver solder? Amazon is a confusing mess.

I don't know these days. The roll I have is an old one from Radio Shack, that I've had for years.

I just googled 'kester silver solder'. You don't need to buy a whole 1 lb roll, there are folks who sell smaller amounts. Maybe something like this:


If you have any other kind of Kester, that should be fine. Like I said, silver isn't critical. What's more important is getting everything down to bare shiny metal, and then cleaning off all of the flux afterward (with Qtips and Goof Off or acetone).
 
So I spent most of today polishing and cleaning. Found some 3% silver solder from the pile and a hot iron and fixed the flyback side cup. Shined the springs and leads and made sure the solder on the springs and diode adhered. Same result. So I dug an HV cage out of another working V2000. Swapped it, and damn it, same result. Here is a video of the problem with the Asteroids Deluxe test screen showing. I appreciate your patience with me on this, Andrew.
 
So I spent most of today polishing and cleaning. Found some 3% silver solder from the pile and a hot iron and fixed the flyback side cup. Shined the springs and leads and made sure the solder on the springs and diode adhered. Same result. So I dug an HV cage out of another working V2000. Swapped it, and damn it, same result. Here is a video of the problem with the Asteroids Deluxe test screen showing. I appreciate your patience with me on this, Andrew.


No prob.

Am I missing something, or is the video just showing a blank screen? (If there is something there, I can't see it in the vid.)

Do you have an HV probe?

Note there may be more than one problem here. (i.e., there could be another issue in addition to HV, causing no picture).

Is the spot killer LED going off when you're in test mode, and does it come back on when you press and hold reset on the game board? (And then go off again when you release reset?)
 
Back
Top Bottom