Repairs, Restorations, Tweaks and Insights
I played Asteroids (the original) in the arcades as a kid. Separate from the arcades, I played the Atari 2600 version of Asteroids a ton more. Asteroids Deluxe, however, is not a game that I played more than a couple of times. Perhaps it was because I was tired of Asteroids when it came out? I don't recall seeing a lot of Asteroids Deluxe machines around.
Regardless, there is no doubt that Asteroids and Asteroids Deluxe are staples of the early arcades. When a local collector offered a completely non-working Asteroids Deluxe (aka dead Asteroids Deluxe) for sale, I decided to grab it. I thought it would be great to add a classic to my gameroom and I was up for the challenge of restoring it back to life.
Let me list a few for you:
It's a vector! I have a soft spot for vector games.
Asteroids gameplay is classic and was copied in various game over the years
Although it's black and white, it uses a blue film to give the game a unique color tint
The gameplay takes place in front of a 3-dimensional backdrop that is achieved through use of a semi-reflective mirror and 3-dimensional cardboard backdrop (which is illuminated by a black light)
While the "music" and sound effects are simple, they are immediately recognizable and representative of early 80's arcade sounds
June 24, 2017
The hardware used by Asteroids Deluxe is very similar to the hardware used by the original Asteroids as well as the game Lunar Lander. Braze Technologies makes a multi-game kit that allows Asteroids Deluxe to play all three!
Braze provides detailed instructions. The process is straightforward and summarized below
The instructions suggest going back and checking your work. You should! When I double checked my work, I found out one of the multi-game boards was installed offet by 1-pin. It's easier to do than you might think because the PCB obscures the socket. A picture of the Asteroids Deluxe PCB with the multi-game kit installed is shown below followed by a picture of the self-test screen. Finally, a picture of Lunar Lander is shown.
June 24, 2017
After installing the multi-game kit, my Asteroids Deluxe received more play over a few days than over the past 6 months. This was, apparently, too much for the game to handle. The 19v2000 vector monitor stopped working and the game was playing blind. This means it was running but you couldn't see anything (but the LEDs worked and you could hear sounds).
At first I wasn't sure whether the multi-game kit had somehow broke and caused the main PCB to not display anything or whether the vector monitor simply wasn't displaying the requested data. I connected my scope to the vector X and Y outputs to determine whether the main PCB was outputting a valid signal. It was, as is shown by the picture below.
Note: The scope I have isn't fast enough to display a vector image. This results is a less detailed image made up of little dots. However, it can approximate the signal and it's clear the game is outputting the lunar surface and other information.
I then started looking at the P299 deflection board. This is the board that takes the signal from the main Asteroids Deluxe PCB and makes the vectors appear on the monitor. It turns out that fuse F101 was blown. This is a common indication that the bridge rectifier DB100 is bad. After removing DB100 from the board, I tested it and it was bad. It was a simple fix to install a new bridge rectifier. After the DB100 fix and a new fuse for F101, Asteroids Deluxe was operational again.
A picture of the P299 deflection board with DB100 remove is shown below.
Very 1980s but very cool. You can see for yourself below.
Very classic and 1980s-ish. The Asteroids Deluxe cabinet, after 99% of my restoration, is pictured below.
The local collector I purchased Asteroids Deluxe from was upfront about the game's condition. He stated it wasn't getting 5V and hadn't checked anything else. When it got into my garage, I wondered what I was in for... I did a quick inventory.
The side artwork was nice. Not perfect, but pretty good
The cabinet was in very good condition. No water damage. No chunks missing
The coin door had been hacked. Somehow the wires were cut and had been twisted back together
The coin door itself was missing some parts and generally looked in poor condition
The inside of the cabinet was exceptionally clean
The control panel looked pretty good although it had what looked like minor cigarette burns
That was a pretty good start. I them moved on to examine the electronics.
Atari A/R looked good. Nothing missing, nothing obviously burnt
The Asteroids Deluxe main PCB looked great
It has a Well-Gardner 19V2000 black and white vector monitor. The phosphor looks good, no obvious burning
Wire hardness was present and looked good
That was a pretty good first impression of a non-working Asteroids Deluxe.
I was eager to see what worked and what didn't. I plugged it in, turned on the switch and.... nothing. No marquee light, no LEDs, no noises, just silence. That was an unimpressive start.
I decided to be safe and disconnected the power going to both the Well-Gardner 19V2000 and the Asteroids Deluxe main PCB. I continued working on the power, or more accurately, the lack of power.
I dug around a bit and realized that Asteroids Deluxe has 2 interlock switches. The game will automatically cut power if either the back door is off or the coin door is open. I used electrical tape and activated the back door interlock. The coin door interlock was already activated since the coin door was closed. I turned on the power again and nothing happened.
After tracing the 120V a little bit through the system, I finally got tired of trying to figure out why there was no AC power. I placed jumpers across the connector on the power brick that checks for the interlock switches. The marquee florescent lamp turned on! It was a little victory.
I would later realize that the interlock switch for the coin door wasn't being pressed in far enough to activate it. In addition to the hacked coin door, it appeared as if a metal piece as missing that fully pressed the interlock. I used a piece of electrical tap to temporarily ensure the coin door interlock was activated. After removing my jumpers, the game would turn on and off normally.
With the monitor and main PCB still disconnected, I decided to check the voltages from the power board. The P5 connector has all the interesting output voltages. Everything looked good. For no good reason, I didn't check the voltage being generated by the A/R board. The A/R board takes the voltages from the power power and creates a regulated 5VDC (among other things) for ICs on the main board. Below is the power board/brick.
I decided to test out the game. I left the monitor unplugged, but plugged in the Asteroids Deluxe main PCB. I turned on the power and WHAM! Immediately the main PCB made a crack sound and smoke started coming out from the main PCB. As quickly as possible, I turned the game off and said a few choice words out loud.
This was the first time I had seen the "magic smoke", also known as a component burning up. I wish I had a video of it since it was a strange activity to observe. The smoke was very dark and directed straight upward. I pulled out the board and looked for damage. CR9 had been destroyed and blew a crevice in the board, taking out a trace in the process. See the picture below (taken after I cut off diode CR9).
The diode that blew was a 1N4001. Luckily these are easily replaced with 1N4004 diodes and I had many of them because they are used in pinball machines. Since the trace was broken, which ran from CR9 to CR8, I needed to remove CR8 when installing a jumper wire. I also replaced CR8 with a 1N4004 diode. The repair is shown below.
(Looking at it from this close-up it does appear I should have used a bit more solder on the "-" end of the CR8 diode. Oh well. Next time I have it out, I'll add a little)
There was no way I was going to plug in the repaired main PCB until I figured out what caused this. I started looking at the Atari A/R board.
The Atari A/R board stands for "Audio/Regulator". This board has the circuitry to drive the speaker and regulate voltage. With the main PCB and monitor still disconnected, I powered on the game and checked the voltage coming from the A/R board. I should have got about 5 volts DC, instead I got 9 volts DC! Yiikes. That's pretty far off.
I initially thought this must have been the cause of CR9 blowing up. After reviewing the Asteroids Deluxe schematics, I couldn't find any correlation between the high 9VDC and the CR9 that blew up. CR9 takes voltage directly from the power board, not from the A/R. Regardless, it was wrong and needed to be fixed.
Voltage regulation problems with the A/R board are common and there is a well known fix. Various video game part stores sell "A/R rebuild kits" which include the common parts that go bad (several capacitors and some other parts). I desoldered the old parts and put the new parts back. I didn't take any pictures of the rebuild process.
Wait! I forgot that I did take one picture. While not included as part of the standard A/R rebuild kits, the resistor R4 on the A/R board got so hot at one point that the outside casing started falling off. R4 is a 100 Ohm resistor. It tested fine even though it had clearly experienced some hot temperatures. I replaced it because it was easy to do.
After rebuilding the A/R, I was pleased to see 5.2VDC being generated from the A/R board. Big improvement! The rebuilt A/R is shown below
With good voltage from the A/R, and previously tested good voltage from the power board, I decided to once again turn on the game with the main PCB installed. This is, of course, the main PCB that previously smoked and I repaired.
I connected to Asteroids Deluxe main PCB and turned on the game. No smoke, but no sounds either. I did notice an LED on the main PCB was illuminated, which I considered a good sign. I watched the main PCB for 30 or so seconds, making sure I was there if it started smoking. Eventually I decided to go look at the control panel. I found the player 1 and player 2 buttons flashing.
I pressed "Player 1" and the game started playing! It was "playing bind" because I didn't have the monitor connected. I pressed the fire button and hear the "pew pew" sound, pressed the "thrust" button and hear rumble of 1980's rocket engines. This was certainly a good step in bringing the game back to life.
With Asteroids Deluxe playing blind, it was time to start looking at the Well-Gardner 19v2000 black and white vector monitor. Prior to this effort, I had done absolutely no work on any monitor (Raster or Vector). One aspect of pinball machines that I like is that they don't have monitors.
Looking more closely at the game playing bind, I saw an LED was illuminated on the monitor PCB (the P299 board). I thought this was a good sign. Perhaps this indicated the monitor was getting power? After discussing with fellow collectors, it turns out this LED is the "spot killer" indicator. It means that something is wrong that would prevent the monitor's electron beam from moving around on the phosphor. If the electron beam doesn't move, it will quickly burn a hole in the phosphor that can never be repaired. One way to look at this is that I should have been happy that the "spot killer" LED was active since it was saving my monitor from phosphor burn.
I pulled the 19v2000's P299 board out from the monitor chassis. It's easy to remove, just unscrew 2 hex screws and disconnect 4 connectors. Just be careful not to hit the yoke or zap yourself (badly). Some would argue I should have discharged the monitor prior to removing P299.
Below is a picture of the 19v2000 P299 board I removed.
My plan was to do the following 4 items:
Examine the board for obvious burns or destroyed components
Check the fuses
Reflow the header pins
Check resistors R100 and R101 (which are known to cause trouble)
I found the PCB to be in very good shape, with the exception of obvious temperature problems around R100. Moving on to check the fuses, I found they were all in good shape. While I was going to reflow the header pins regardless of what I found, it was obvious some of the pins had cracked solder joints. A couple examples of this are shown in the picture below. At least 4 of the P500 pins have cracks. These appear a dark circle-like shapes inset within the base of the solder circle.
Reflowing the connections solved the cracked solder issues. The last item I planned to do was to check resistors R100 and R101. It turns out that R101 was fine but R100 was not working at all. It was an open circuit. R100 was the resistor with the burn marks below it. I learned that R100 and R101 have a long history of causing trouble. Atari issued a service bulletin to arcade operators that stated both R100 and R101 should be removed and replaced with jumpers.
Based on this information, I removed R100 and R101 and replaced them with jumpers. It was recommended that the jumpers use 18 AWG. I didn't have any 18 AWG handy, so I decided to run two 20 AWG wires for each jumper. A picture of resistors R100 and R101 is shown below, immediately followed by a picture of the newly added jumpers.
Having taken my initial stab at repairing the Well-Gardner P299 board, it was time to test it out.
I put the P299 board back into the monitor cage, connected the wires, and turned on the game. I didn't see anything right away, but that's common because the monitors need to warm up. After a few seconds, did I see something? I did! I could see the Asteroids player ship and several asteroids!
But wait... all was not well. The image was noticeably dimmer in the center of the screen than towards the outside. Also, the screen was bouncing between two size: too big and even bigger. I attempted adjusting the X and Y linearity pots to see if I could fix the screen size, or to determine whether the pots were bad and causing the screen jumping. The screen jumping had no correlation to tapping on the linearity pots and no adjustment would correct the size.
This was another big step in bringing Asteroids Deluxe back to life, but clearly wasn't the end state. More work was needed.
I learned from my fellow collectors that monitor blooming is often caused by a bad or corroded high voltage diode. This high voltage diode is connected inline between the anode and the flyback. As the name would imply, there is some serious (life threatening) voltage going here and one must approach it with care and caution. In a nutshell, this meant I needed to learn how to discharge a monitor.
WARNING: This section is in no way a lesson in monitor discharging. This can kill you. Please find some official instructions before attempting this potentially life threatening procedure. It is probably best if you get a professional tool, such as a HV probe.
I was a bit worried about performing this for the first time. In retrospect, it was probably good that I worried about it because I did a lot of research into the process. For example, it is easy to miss the fact that discharging a B&W vector is slightly different than discarding other monitors.
I performed the following steps to build my discharge "tool" and discharge the monitor:
Found a long flathead screwdriver with insulated handle
Purchased some large alligator clips
Purchased a 1M Ohm, 3 Watt resistor
Found some 14 gauge wire, make it into a long piece and a short piece
Soldered one end of the long wire to an alligator chip
Soldered the other end of the long wire to the 1M Ohm resistor
Soldered the opposite end of the 1M Ohm resistor to a small piece of wire
Soldered the unused end of the small wire to the remaining alligator clip
Clipped the alligator clip on the long wire to the screwdriver blade
At this point I had a circuit with a limiting resistor that runs from the screwdriver blade to the alligator clip. The process of discharging a monitor involves shorting the energy in the monitor from the anode to the monitor's frame. To achieve this, I did the following
Connected the available alligator clip to the monitor's frame
Placed my hand in my back pocket
Carefully inserted the tip of the screwdriver under the anode cup until it touched the anode.
Waited 5 minutes, repeated these steps again
If the resistor isn't in the circuit, a nice pop sound that tells you the discharge has occurred (so I have read). With the resistor, you don't get any feedback. This is a bit worrisome because you never know if you did it right. If is why it's probably best to have an HV probe and a DMM since it would tell you the voltage remaining.
The picture below shows the "tool" I made. The resistor is inside the back plastic casing.
With the monitor discharged, I started looking into the HV diode.
With the monitor in a safe (discharged) condition, I opened up the high voltage cage and saw the existing HV diode. The boots holding the HV diode were in sad shape. When I moved one of the wires, the HV diode fell out of the boots (apparently not) holding it in place. The HV diode's leads were nowhere in site. Apparently the diode was barely resting in the boots and barely making any electrical connection. Below is a picture of the HV diode boots, followed by the HV diode. I suppose it's amazing the monitor was working at all.
I was lucky that a local collector had a spare HV diode and I was able to pick it up rather than ordering it and waiting for it to arrive.
At first I was going to attempt to mount the new HV diode in the existing boots. After attempting to clean the corrosion from the boots and finding that even more plastic from the boots cracked off, I quickly decided this approach wasn't going to work. I was going to need to solder the new HV diode directly in-line.
I cut off the boots and stripped the anode wire and the flyback wire. There isn't a lot of excess wire on the flyback, so I was very careful to not accidentally cut off the wire while stripping it. Even being careful, the odd angle and strange insulation on the flyback caused me to cut off rather than strip. The flyback wire was now 1/4" shorter. My second attempt was successful. The two stripped wires are shown below.
I soldered the HV diode onto the stripped wires. The "-" side of the HV diode went toward the anode/CRT, which was the red wire in my case. It was important to get the diode's direction correct for two reasons. First, and perhaps obviously, the monitor wouldn't work if was connected wrong. Secondly, I was going to cover the exposed leads with RTV silicone as an insulator and it would have been a pain to get the dried RTV silicone off and reverse the diode's direction.
It was hard to apply the RTV silicone in aesthetically pleasing manner, so the result is certainly ugly. The goal was to insulate, not create art. I was focused on ensuring solid coverage for the insulation. Below is a (poor) picture of the result.
The RTV silicone takes about 24 hours to dry, so I needed to wait to try it out!
I carried the monitor over to the Asteroids Deluxe cabinet and connected it to the cabinet. It's nice that you don't need to mount the monitor in the cabinet to test it. The monitor was sitting on the ground behind the cabinet. I screwed the high voltage cage cover back on, plugged the game into AC power and got ready to flip the power switch.
Within a few seconds of powering on the game, there was a beautiful Asteroids Deluxe game image displayed! The vector lines were solid, straight, bright and sharp as a tack. Success! :)
Below is one of the first images I saw after bringing the game back to life.
(Yes, the image is reversed. This is because the monitor's image is reflected off a mirror when the game is in the cabinet!)
Here is an image of Asteroids Deluxe self test, which does a good job of showing how nice the vector monitor has turned out.
The coin door was pretty grubby. It looked worse than is apparent from the image below. It had a fair amount of rust.
I sanded off all the rust and paint that was falling off. A picture of this process is shown below.
The painting stage is shown below. This is after a couple coats of "Hammered" black.
The "hammered" is followed by a coat or two of black semi gloss. Below is the 99% finished product. I'm still missing the flaps for the coin return, but it looks a zillion times better than before!
The operator that originally owned this game, or someone in the games history, cut several of the coin door wires. This caused the coin door lights to not work, the coin sensors to not work, and the slam tilt to not work. It was fairly easy to repair the cut wires since they where near each other and had matching colors. I stripped the wires, twisted them together, added some solder, and put electrical tape over the connections. Using "shrink tubing" would have been a more professional option, but I didn't have any when doing the repair.
The picture below shows in coin door wiring. I've circled (in yellow) the places that were cut and repaired.
The counters in the coin door show a fairly small number of quarters were placed into my Asteroids Deluxe while it was earning for an operator. One coin chute registered 9,413 quarters and the other registered 7,131. That's a grand total of 16,544 quarters or $4,136. Not to shabby I suppose.
It's not fair that I've only got an after picture, but since I didn't take a "before" picture this is all you get. I'm very happy with how the control panel cleaned up. I thought it had a couple of spots with cigarette discoloration (but not burns), but I was able to completely clean it up using "magic eraser". There was a small area on the front where the black was scrapped off, leaving some white speckles. I was able to cover that up with a black sharpie. The cleaned up control panel is shown below.
Asteroid Deluxe's speaker is just below the marquee. The same mechanism supporting the speaker also holds two florescent bulbs. On bulb is for the marquee and the other is for the 3D artwork. While the first bulb is a standard florescent, the 3D artwork bulb is a "black light white" bulb. The structure that supports the lamps and the speaker is shown below.
While the marquee lamp worked fine, the black light lamp didn't. I realized that someone had removed the starter for the blacklight lamp. When I put a starter in for the blacklight, it became apparent why the starter had been removed. The black light took forever to light (or never did). The starter just kept clicking (and a little spark could be seen through the hole on the top of the starter). It became clear I needed a new lamp.
It's easy to purchase the wrong blacklight for Asteroids Deluxe. Most stores will sell a blacklight that is technically "black light blue" or BLB. This is a florescent bulb that looks like it has a black coating (or a very dark blue coating). This is not what should be used in Asteroids Deluxe. Rather, asteroids deluxe expects a "blacklight white" or BL lamp. While these are available all over the internet, they aren't as easy to find locally (at least for me).
The GE 35884-F15T8/BL appears to be a good match. The appropriate starter for both lamps is an FS2.
Below is a picture showing the two starters. More correctly, one starter and one connector for a missing starter.
While Asteroids Deluxe should store the top 3 high scores, my Asteroids Deluxe was pretty inconsistent in this area. It seemed that sometimes the high scores just went away. Other times the top high score went away and the 2nd or 3rd high score was promoted to #1. It was pretty strange behavior.
I looked a look at the schematic and saw there were several items involved in making the high scores work. There were a couple of caps that ensure -28V is available to the EAROM chip, the EAROM chip itself (2055/7055) and some ICs to handle the logic. Without checking anything, I ordered a 2 new 1uf 50V radial caps and swapped them. I initially though the problem was fixed, but realized after a couple additional power cycles that the original problem was back.
At this point I decided I should check the -28V going to the EAROM chip. It turned out to be -28.5V, which is just fine. However, I was a little careless when checking this voltage and briefly shorted this pin with one of the pins next to it. This caused something new to break. Now the high scores would typically display values like "FAFAFAFA" or "FBFBFBFB" when the game started and then, like the game realized they were invalid, they would disappear.
My first though was that the 2055/7055 chip was going bad and I'd just pushed it over the cliff. I ordered a new 2055/7055 and installed it. This chip is in a socket, so it's quite easy. This made no behavior change.
There only 2 other chips that were likely to being causing trouble: a 74175 chip and a 74374 chip. I went to a local electronics retailer and found they had the 74175 in stock, but not the 74374. I figured at least I'd be able to replace one of them. The 74175 chip is just above the socketed 2055/7055 chip. The picture below shows the newly installed 74175. It's the one that says "NTE" upside down.
After replacing the 74175, the high scores started working perfectly! My best guess is that the initial problem was really a bad 2055/7055 EAROM. However, I probably fried the 74175 when checking the voltage, creating a secondary problem. All's well that ends well.
I was a little surprised that after replacing the 74175, I got the high scores from Asteroids Deluxe play the previous evening! I believe this is because I didn't play and games between the previous evening and the repair. This tends to show that the high score storing functionality was working correctly, but the ability to read the high scores back was broken. When I fixed the 74175, the ability to read was restored.
Ever wondered how you access the micro switches or the leaf switches? The Asteroids Deluxe control panel has two bolts with wing nuts that hold the control panel down. You need to reach up behind the coin door and remove the wing nuts. Once done, the control panel will tilt outward because it is on hinges. You can see this in the picture below.
Below are a couple of pictures of the Asteroids Deluxe main PCB.
The original software in ROM for Asteroids Deluxe was deemed to difficult by many. Atari responded and released an updated which allowed the difficulty to be changed. The title of the document describing the changes, as well as the installation, was "New Program with Easier Game play for Asteroids Deluxe". I've duplicated some information from this document below.
The technical manual describes the game play correctly if the game PCB option switch at R5 is set to "hard." If you set the switch to "easy", then the following game-play changes happen:
Four large asteroids begin the game. The second wave of asteroids begins with five, and the subsequent waves start with six through nine large asteroids. In addition, the asteroids move much more slowing across the screen. (If the operation switch is set to hard, the waves begin with six to nine large asteroids.)
The large ships ("death stars") when shot will break up into three slowly-moving diamonds. (If the option switch is set to hard, diamonds would immediately begin chasing the player's spaceship at high speed.)
The game uses 5 roms. Four of the 5 are close together and an image of these 4 is below. While these chips have stickers that appear to indicate the original (hard) version, they actually contain the newer version allowing for easier or hard gameplay. These chips are PROMs and have been reprogrammed with the new code. Since these chips can be around $10 (USD) to purchase, having the newer code saved me about $50 (USD).
Pictures that haven't been added to a section yet...