Repairing Wells Gardner's K6100 Color Vector Monitor
By, William Stephens
(clicking any picture in this article will bring up an enlarged version)
A customer of ours has been after his neighbor to sell him a 1983 Atari Star Wars videogame over the past few years. His neighbor would not sell... he liked it too! One day as our customer pulled up to his place of business he noticed the Star Wars game out back by the dumpster having been thrown out. Well... he snatched it up in a flash, took a cursory look at it and realized why they had thrown it out.
The cabinet was in structural distress (falling apart) and the neck was broken off the monitors picture tube. He gave us a call and we came and rescued the vintage piece. We fixed the cabinet and re-geared the controller. The circuit boards and power supply were filthy and had some flakey connections but other than that they were in fine shape. The monitor however would take some work for it to ever display "May The Force Be With You" again.
This picture shows the finished and restored monitor mounted in the Star Wars Cabinet, but when it came into the shop, it looked like a pile of junk that was incredibly dirty. The first thing I did was to take a bit of an inventory to see if everything looked like it was there. In general the answer was yes with the exception of the faraday shield that covers the high voltage unit. This cover is usually missing to aid in cooling the high voltage components, especially the rare and expensive high voltage transformer. I think this cover should be kept on the H.V. unit to keep things from falling into this voliatle circuitry. I plan to make a new one with some fans in it.
At this point I took the monitor and put it on the "waiting for parts" shelf and placed an order with Richardson Electronics for a 19VLUP22 (19" 100 degree) color picture tube. Although they rarely have the tube I want in stock, the do get them usually in a week or two. It's time to get on with other business and wait for the UPS man to arrive with the precious cargo.
About a week later I pulled the monitor off the shelf and removed the chassis from the monitor frame for cleaning. I also set to work on the old picture tube to safely remove the yoke and purity ring assembly. Seeing that the tube was already broken I opted to nibble away at the glass with some needle nose pliers to remove the rings and yoke. Over the years these parts have welded themselves to the cloth friction tape on the tube neck. Prying and twisting these things free can often result in breaking these embrittled plastic parts, once that happens you're kind of screwed. The slow methodical nibbling of the broken glass of the neck put no stress on these parts which eventually slipped free and fell from the tube in perfect condition. I carefully cleaned the yoke and set it away for safe keeping. The dynamic convergence ring assembly on the other hand had to have all that nasty glue removed. I decided on disassembling the ring assembly which makes cleaning much easier... BE CAREFUL!! you can still break these fragile plastic parts.
With those two critical tasks complete I turned to the chassis which I completely disassembled and then gave each of the parts a good washing. The chassis on the K6100 is aluminum and seeing that I am going to replace all the driver transistors when I re-cap the monitor, I removed the transistors and sockets and used some Scotchbrite on the chassis to really clean and brighten it.
While disassembling the chassis the G2 and Focus grid wires from the H.V. unit going to the neck board broke at the interconnect on the H.V. unit. These wires, (544volts and 5,700volts respectively) are special in construction having thicker than normal insulation which makes crimped connection prone to failure. I manually re-crimped the connections and put in a new interconnect socket to fix this small problem. I also noticed that the X axis yoke fuse was blown, but without the tube installed yet, it will be a little hard to discover why.
I went ahead and replaced all the driver transistors and some electrolytic capacitors mainly in the power supply board. The deflection board does not have many capacitors as would be found in a raster scan monitor... then again a raster scan would not have 6 driver transistors in complimentary pairs. It's all now waiting on the arrival of the picture tube.
About two weeks after I ordered it, the picture tube from Richardson Electronics arrived. By this time most of the refurbishing, re-capping and cleaning had taken place, so at this point it is left up to basic assembly. I mounted the hardware to the picture tube, installed the chassis, loosely placed the yoke and convergence assembly on the neck of the picture tube, hooked up all the plugs and plugged it into the wiring harness of the Star Wars game cabinet.
Upon power-up, a buzzing came from the monitor and the image was "trippy" to say the least... kind of a color bloom about the center with a big white "squark". I also noticed a glowing inside of the picture tube socket... I quickly disconnected the power.
The spark gap on pin 7 of the picture tube was causing the glowing in the socket because I stupidly wired the Focus and G2 grids backwards when I repaired the interconnect at the H.V. unit case. A quick swap of the wires in the interconnect solves that problem. It is easy to make this mistake with this monitor because the Focus wire coming off the H.V. adjustment block is red and the G2 is black. At the interconnect the Focus wire becomes white and the G2 grid wire becomes red... confusing??? I thought so.
I plugged the monitor back into the wiring harness and this time there is no glowing, buzzing or light show... as the tube warmed up I could see the "Star Wars" logo slowly scrolling up the screen in bright blue and other features showed up in their correct colors. The convergence was still way off and the purity was not all that pure but I had not converged the tube yet. I thought the results looked very good until I noticed that the spot killer LED was glowing. I also noticed that areas of the screen were dropping out.
I visually checked out the spot killer circuit and noticed that I had replaced 4-.47uf tantalum capacitors with 4-1uf electrolytic capacitors (C800-C803) from the re-cap kit. The 1uf value was fine, I just hoped that I had not put one of them in backwards... I didn't. This circuit is real simple and has only 3 active parts, one of which is just the LED driver. A quick check of voltages in the operating circuit showed that Q801 was intermittent under load while it tests OK on a tester. This caused the voltage divider circuit quickly to flip between balance and unbalance causing the beams to flicker on and off and make the LED to appear to be on when in reality it's just blinking so fast as to look continuous. I replaced that transistor and the spot killer shutoff on queue and the dropouts in the screen stopped.
Now it's time for the fun part, converging the picture tube. I set the 3 magnet pairs opposite each other nulling out the their magnetic influence. I then placed the ring assembly so that the first (convergence) pair was over the G3 gap in the picture tube neck, I locked down the ring assembly to the neck. Next I slid the yoke all the way back to the ring assembly and turned on the monitor. I set the game for test mode and used the red grid setup screen to do the static convergence (placing the yoke). With the yoke slid back the screen displayed the grid a bit shrunken and a specturm pattern of color blue to red to green across the picture tube face. I slid the yoke forward slowly until the grid was as completely red as I could get it. I used the rubber wedges from the old tube (save everything) to hold the yoke in place and then tightened up the yoke clamp screw. I put a few dabs of silicone adhesive on the wedges to hold them in place and that completes the static convergence.
Dynamic convergence is a bit different and goes something like this... The first magnet ring assembly is the convergence magnets. I rotate, split and adjust those two to perfect the pure red field I created by the static convergence. Next I put up a busy screen with some writing both top and bottom so I can see the overlap of the three gun colors. I tackle the middle or second set of magnet rings and adjust their position so the red and blue images line up (this can be a bit fiddley). Once I have this optimim, I go to the third set of magnetic rings and move the green image to align with the red/blue ones. Once I have this optimim I go back and tweak the positions for overall alignment of the colors. A note here is important because you can usually never get this alignment perfect. This is a compromise situation where you try to achieve perfection but you will fall a bit short... don't take this failure personally... everybody compromises.
Once things seem to be adjusted correctly you may want to adjust the RGB drive and bias controls, these are usually set fine but you may want to fiddle with them. Lastly, the Focus and brightness should be adjusted for peak performance, some silicon adhesive is applied to the ring assembly and that's it!
I did modify this monitor by making a new faraday shield for the H.V. unit with two 12 volt fans installed for cooling. I wired the fans in series so they could run from the 25.7 volt low power supply inside of the H.V. unit cage. There are 2 soldered up holes on the circuitboard by the transistor connector (red) where you can tap this 25.7 volts... a nice clean factory looking modification.
As a last note: I never figured out why the X axis yoke fuse was blown, once replaced it never blew again.