Hello, newsgroup regulars and non-regulars alike -- I thought I might share some of the things I've learned over the last week and a half while attempting to repair my Vectrex machine. I'm not an expert in electronics, but I know a few things and this is one of those occasions where I actually diagnosed the problem correctly and fixed it! If you're a seasoned repair tech, this message will probably bore you from here on out and you'll wince at some of my terminology. But if you're like me and enjoy tinkering, you may also enjoy my notes below.

------------------------------------------------

First of all, my machine is a Japanese Bandai Vectrex. It has Japanese writing on all of the console and controller decals. I don't know if it's any different internally from the American Vectrex, since the schematics I have seemed to match up pretty close. Exceptions are noted below.

The symptoms: The monitor on this machine used to work just fine with a nice picture until a couple of years ago. Then suddenly, the machine would work upon power-up for about five minutes, and then the monitor would go crazy, with vectors jumping around and everything. Shortly after this, The picture would fade out slowly. As the picture faded out, the image would increase in size -- sort of a 'zoom in, fade out', if you will. After taking the back cover off of the Vectrex the first time, I noticed that the tube filament (the orange glow you see in the neck of the tube) would fade out along with screen. After this the monitor would be dead until you turned it off and on again. The longer you waited with the power off, the longer the picture would stay stable before flipping out the next time. During monitor freak-out periods, the game itself would play fine, as evident from the game sounds I was still hearing.

My first thought: Oh shit, it's the flyback transformer. Probably completely irreplaceable in a unit this old and this obscure. A nine-inch vector monitor isn't exactly state of the art anymore. Not that it was to begin with. :)

My second thought: The monitor DOES work for a short time, and in my experience when flybacks go, they GO. So something must be overheating. I took the machine apart and blew an entire can of cleaner/degreaser on the board parts that I could get to. I was sure that removing the thick layer of dust from the logic board would help. But this didn't fix the problem.

My third thought: This machine is NOT built for service. I spent the better part of a day figuring out how the two boards were affixed into the plastic shell. I took careful notes on which wires I had to desolder and where molex plugs went.

For the record, here's the order in which I took apart the Vectrex. There's probably easier ways to get to all the board components, but this is how I did it:

1. Turn the machine down on its face, and remove the back cover. This just lifts off. Probably the easiest part of the repair. :)

2. Unsolder the four monitor yoke wires from the power board. The power board is the vertical board in the left of the cabinet. These wires are in a small line in the center of the board and are colored blue, red, green, and yellow. The four wires loop around and attach directly to the picture tube. I do not understand why the yoke wires are not attached to a molex harness. All of the other monitors I've worked on in video games have a harness that is easily removable. I intend to go back in when I have time and install a harness. Maybe the American Vectrex is different?

3. Unsolder the three wires in the bottom front of the power board. I think these come directly from the transformer at the bottom of the console. There is one white and two red wires here. Make sure to note which red wire goes where. I put a piece of scotch tape on one of them to differentiate them. Also unsolder the short ground braid that connects the two boards.

4. Unplug all molex harnesses between the two boards and remember (or write down) where they go.

5. Clip the white wire tie that holds the neck board to the picture tube. Remove it, and then *gently* pull the neck board off of the picture tube, straight back. Be careful -- I once dropped the back door of a Phoenix machine into the cabinet and sheared the stem of the picture tube clean off. I got lucky -- the tube didn't shatter into a million pieces because it was such a clean break.

6. Get a flathead screwdriver and attach a wire to the metal part of it. Then attach the other end of the wire to something grounded. I used the metal frame of the monitor. There's not much metal inside of a Vectrex, but this seemed to work okay. Then use the screwdriver to remove the anode cap (the suction cup looking thing) off of the picture tube. It has two prongs which need to be squeezed together to unhook it. You may hear a pop when you first put the screwdriver under the rubber. This is normal -- you're discharging the picture tube.

7. Now unscrew the green ground wire from the upper left corner of the picture tube. Also unscrew the two screws holding the power board to the plastic carriage. The power board should come out completely now. Sit it aside and take out the two big screws holding the carriage to the cabinet. Sit this aside too.

8. Now the fun part. Pull the on/off knob straight off. It's hard to do, but it should come off completely. Now take out the one screw that holds the cartridge port and the logic board together. Also remove the four big screws that hold the logic board chassis to the cabinet. Two of these are above the transformer. The entire chassis should come out now and the cartridge port sheath should come off entirely. Put the picture tube in a safe place where it's not going to fall. I put mine next to my cats food and then moved it when I found out that kitty wanted to play with the Vectrex too. Bad kitty, bad!

9. I could never get the logic board completely separate from plastic, but I did get it to the point where I could solder on the rear side without melting anything important. Turn the board upside down and unsolder the two reset button wires from the bottom of the board. There is one gray and one black wire. Then turn the board back over and take out the last three screws. You should now be able to flip the board by itself over to work on it. Whew!

Things I replaced: I've installed many, many Zanen cap kits over the years, so I figured the one thing I would do for sure was replace all the electrolytic caps on both boards. There are four of them on the logic board, and 22 on the power board. After looking at the schematic and talking to a few repair-head friends, I decided that the monitor filament circuit was suspect as well. I got a couple of suggestions to replace the big heater resistor in the circuit. This is R529, a 1W-3.3ohm carbon-film resistor. My board had some very small, almost nonexistent burn marks on the bottom side of the board under the resistor. This didn't alarm me too much, as hot components will do this, even if they're not bad.

I also got a lot of really great suggestions from another Vectrex-head (unfortunately his name and E-mail are on my work account, or I'd give him credit here) who immediately suspected transistor Q502. I didn't think that transistors could overheat -- I thought that like flybacks, when they died, they died. But I decided to replace it as well to remove it from the equation.

I found all the parts I needed at the local Tinkertronics, with a few exceptions:

1) The 3.3ohm-1W resistor was just not going to happen. I managed to rustle up a 2W-3.3ohm resistor, though.

2) Q502 is a BU407 NPN transistor in the 220 package (square with 3 leads sticking out of the bottom). I had to get an NTE part for this one, and for reference, the NTE equivalent is NTE 379.

3) I didn't find several electrolytics, including C409. Note that it's okay to go over on the voltage of the replacement caps if you need to. I had to do this for a few of them. For example, the three 1000uf-25V caps were replaced with 1000uf-35V caps.

4) I ordered C409 and the rest of the caps I needed from Zanen Electronics. Zanen did NOT have the 1W resistor I needed, and did NOT have C409. For reference this is a 0.47uf-50V nonpolar electrolytic capacitor. It's the only electrolytic cap in the machine that's nonpolar. Kind of obscure. But the nice folks at Zanen said that they could make one from others. Sure enough, when I got the cap, it was really two caps with their leads soldered together. I was impressed! Those guys at Zanen really know what they're doing.

One other important point: I found out after I'd placed my order that Zanen actually sells cap kits for the Vectrex! I wish I'd thought to ask them before I started scrounging around for parts. Oh well. I buy cap kits all the time from them for videogame monitors. I've said it before and I'll say it now: replacing caps works wonders on a sick monitor.

I discovered a couple of inconsistencies between the schematic's part list and my machine. These caps may have originally matched the manual, and later replaced, though I can't tell for sure:

1) C507 is 47uf-100V, not 4.7uf-100V.
2) C508 is 47uf-100V, not 15uf-63V.
3) C516 is 3.3uf-350V, not 3.3uf-50V. Going over on voltage a little is one thing, but this is quite a bit.
4) C228 on the logic board is nonexistent in my machine. The board is labeled, but no part has ever been there.

Once I had all my parts, I spent about four hours installing them and tightening up solder joints. I didn't see any particular joint that stood out as definitely cold, but there were some where the solder was dull. Some of the joints looked like they didn't get much solder at the factory. I remedied all of these and made all the joints on both boards nice and shiny.

One thing to note on the power board is that two of the caps I had to replace (including C409) live underneath the large heat sink in the middle. I will reiterate: The Vectrex was NOT built with service in mind. To remove this heat sink:

1) Unscrew the two screws on the top side of the sink that hold it to the IC.
2) Unscrew the two transistors that sit on one side of it, one screw each.
3) Unscrew the six screws on the bottom side of the board. Pull the heat sink out gently and sit it aside.

While the heat sink was off, I cleaned off the old heat-sink compound and replaced the mica on the two side transistors. When I put the heat sink back on, I put some new heat-sink compound on the two transistors and the IC. It was the least I could do to make this board prettier. :)

I almost replaced the +5 and -5 voltage regulators, but decided against it. These are the two transistors on the side of the large heat sink on the power board (a 7805 and a 7905, respectively). I decided against it because I figured that these probably powered the logic board, which by all accounts was working fine.

One of the Vectrex-heads that I talked to mentioned that when he replaced the two caps under the heat sink, he put them on the BACK side of the board so that he could replace them more easily in the future. I realized that I hadn't done this when I was done swapping out parts, but I probably should have taken his advice. Oh well. The heat sink fit back on just fine, though.

After swapping out 4 caps on the logic board, 22 caps on the power board, Q502, and R529, I was ready to put the machine back together and see if it worked! Before this, though, I blew another can of cleaner/degreaser on the boards. Most of this went to the logic board, which will always be dirtier than the power board since it sits horizontally. Then I carefully put the machine back together, following my notes in reverse order.

Allow me to say it again: I'm really glad I took good notes while disassembling this beast. Other monitors are easy to pull apart, but this one is a real BITCH. Be warned.

I left the back cover off of the unit so I could watch in case anything blew up when I turned it on. At least that might give me a clue about where to look in the event of failure. I plugged it in, turned it on ... and the familiar Vectrex "da-da-dummmm" came on. I hadn't broken it! And I noticed that the background hum on the speaker was quite a bit softer. The work of new filter caps, no doubt.

I played Mine Storm for about thirty minutes before I realized that I had played Mine Storm for thirty straight minutes without the monitor going nuts. It's fixed!

The problem was therefore one (or a combination) of three things:

1) A cold solder joint with an intermittent connection. I don't think this is very likely, but possible.
2) A dried-up or nearly dried-up electrolytic cap. Likely.
3) A thermal problem with a key monitor component, either Q502 or R529. Also likely.

And after working on somewhere between fifty and a hundred monitors, I still have only seen one bad flyback. It was on the dearest item in my game collection -- my Robotron -- and it was shooting blue sparks out of it like a Frankenstein movie.

I still need to figure out what all the adjustment pots on both boards do, because the alignment and size of some vectors suggest that there's an adjustment out of whack somewhere. If someone out there actually managed to read this far, and knows how to align the screen, could you send me an Email? I have Sean Kelly's excellent multicart, which has a great test pattern generator on it. Thanks for reading! I hope my rambling was helpful to any would-be techies out there. :)

Addendum:

  In regards to "that suction cup", just make sure as hell you have discharged the tube before you go putting yourself at risk of getting a 20,000 volt shock.  Not a good way to end the day, or your life.
 
  -Jeff