_______________________________________________________________
3/12/04 Repairing a SpaceFirebird PCB
Thanks to Gary Wilkinson
I picked up a faulty Hoei Spacebird cocktail machine off e-bay for a few pounds. When I got it home and powered it up I found the game was actually Nintendo Space Firebird. The game booted and played ok except for some very annoying graphics errors. Some of the spacecraft would turn into characters and the score and text accross the top was corrupt displaying random characters. Another fault I discovered was that after 10-15 mins of playing the display would suddenly develop a red hue distorting all the colours. After some digging around I found out that the Hoei Spacebird was basically a rip off of Nintendo's Space Firebird. The PCB's artwork was copied almost exactly and after finding ROM images on the internet for both the Hoei and Nintendo versions discovered that some of the Hoei ROM's were identical to the Nintendo ROM's. After running these ROM images in Mame it was clear that the Hoei clone was the Nintendo version with just the "(C) Nintendo" removed from the game program. Now down to the repair. I read the ROM images that were in my machine and found that they were actually the Nintendo version and not the original Hoei version. The manufacture date on the 2716 EPROM's was 1987 whereas the rest of the machine was 1980, so the ROM's must have been replaced with Nintendo ROM's at some point. I then programmed a new set of EPROM's with the Hoei images to rule out any incompatibility with the Nintendo ROM's. The corrupted graphics were still there. As these were new EPROM's it also ruled out faulty ROM's. I then proceeded to look at the Video board. I tried swapping the D5101LC 1024bit SRAM chips around and found that this caused differing graphics problems making the game worse. I therefore concluded that 1 or more of these must be faulty. After being quoted £17 ($34) per chip for replacements I looked on E-bay and found a shop selling a faulty CPU and Video board for the Nintendo Space Firebird. Upon receipt of these faulty Nintendo boards it was clear how close the Hoei copy is to the Nintendo parts. Some of the parts were slightly different values e.g. 3K3 instead of 2K2 resistors and 68nF caps fitted instead of 100nF caps and some of the IC's were different but equivalent. The CPU board from the Nintendo was badly damaged with the corner where the Z80 should sit smashed off and there were missing logic chips on the Video board. Luckily though all the RAM, SRAM, PROM and ROM chips were still fitted. I removed the 8 SRAM chips and fitted them into my Hoei video board and powered the machine up. This fixed all the graphics issues but the red hue still appears after a few minutes playing. I guessed that this problem may be caused by something overheating so I felt around the board and discovered that IC 3i (SN74S201N) was getting very hot. I replaced this with one off the Nintendo board and this cured the red problem. With all graphic problems fixed and the red video problem resolved I thought it was fixed. However, I soon discovered that I had 1 more problem. My Space Firebird now has a very odd fault. All the "projectiles" have vanished. The game works perfectly and plays correctly except that you cannot see either the missiles fired by your craft or the missiles and bombs fired from the alien craft. All the other graphics are there including the background stars. I was under the understanding that the same circuit would generate all the moving graphic entities or sprites. Surely if all the space craft exist then surely the missiles would to???? I have tried swapping all the RAM chips around to see if I can create a differing fault, but with no joy. I have tried replacing all the PROM chips, the larger SRAM chips and the ROMs from the Nintendo board but the missing missile anomaly is still present. Have you ever come accross this type of error before? Any ideas?

_______________________________________________________________
3/12/04 Repairing a DigDug PCB
The board would play but the graphics were messed up. It looked like every other pixel for the characters was shifted down about an inch so it would look like black bars running around the screen below the characters. The screen also had alot of sparkles throughout. I figured that a data bit was out but wasn't sure where. I first eliminated the background (playfield) graphics by removing the ROMs at 4F and 4J. Now I could clearly see the messed up character graphics. Looking at the schematics, the characters are controlled by the motion object circuits. Poking around with the scope, I noticed that signal M0 looked like it was riding too high (all of the signal was too close to +5V, never going to ground). I followed this back to IC 8D (74LS298) and found pin 15 (signal M0) was too high and the inputs at pins 1 and 2 looked ok. Changing this chip fixed the graphics problem. I then proceeded to clean the pins on the socketed chips and promptly broke off a pin off of the custom IC at 2L/M (a 0894). I replaced this with a 0859 from a dead Galaga PCB. Now the game works fine.

_______________________________________________________________
7/9/03 Repairing a Raiden II PCB
This board was from a friend of mine. It played fine but the sound had a massivly noisy hum/squeal. Kind of like EEEEEEEEEEE! OK I cant type what I heard. Poking around with the scope I found that on pin 10 and 11 (which are tied together) of the hybrid chip (the SIPP package that is potted and has components within) in the sound circuit had a sawtooth wave on it. And it matched the sound I was hearing. I did not have a replacement chip but I found that if I cut pin 10 away from the PCB, the hum was gone and the sound level was OK. So it is sort of repaired for now.

_______________________________________________________________
1/26/02 Repairing a Centipede upright
I bought this at the January 26th auction from the dead row (back row). It was missing the monitor and monitor glass. I also won an auction for the monitor glass and an ARII. After the end of the auction we had some time to kill so I broke out the tools and a meter. First tried to power it up but got nothing. Then checked the power outputs and found I had AC voltages but no +5VDC. I then tried the spare ARII with no change. I then pulled up the power brick and found that a wire had fried off of the bridge rectifier. Cleaning it as best I could, I reconnected it and tried out the game. We now had +5V and the game played blind! but oddly the +5V LED on the game PCB did not light. I reinstalled the original ARII and the game still worked fine minus the +5V LED. So don't always trust that LED on the game PCB! I brought the game home, replaced the bad wire on the rectifier, installed a monitor, installed a trackball rebuild kit and was up 100%!

_______________________________________________________________
10/14/01 Repairing a Qix with no sound
My neighbor Gary bought a Qix game from the Mesquite auction yesterday and we found that both speakers were disconnected. We hooked them back up but still got no sound, just some pops and clicks. Gary had some extra Qix PCBs so I swapped the sound PCBs but there was no change. I then swapped the ROM U27 which is for sound generation, this ROM is on the ROM PCB, not the sound PCB. This did not change anything. The odd thing about this layout is that there is no other connections to the ROM PCB, all traces go to the Sound PCB via a ribbon cable (J5). I then replaced the ribbon cable and bingo we now had sound! I tried a collection of these cables and only found 1 out of 6 to work! I tried these cables in other positions on the game and this only caused Qix to not come up at all as expected. So check these cables carefully, your Qix PCBs may not be bad, just the cables are toast.

_______________________________________________________________
6/24/01 Repairing a Vball PCB
This is a PCB I bought as a lot of PCBs from the auction. The right hand side of the graphics had vertical black bars running down the screen. I poked around the data/address lines in the video section and found two without any signal. Upon closer inspection, I found a couple of caps that had been squashed and there legs were shorted. Removing the shorts fixed the PCB.

_______________________________________________________________
6/16/01 Repairing a Starcastle
This is my third Starcastle game I have purchased from the auction. Heck I am determined to get one working. Well this was my cheapest one and easiest to fix! The game had a stripped/rusted game PCB with no sound PCB. I haven't tested it yet but I have plenty of spares PCBs (4). So I dropped a PCB in and found the monitor was dim and had little vertical deflection. I disconnected both the horizontal and vertical output transistors and got a perfect 2"x2" screen! All you will se is the mother ship taking up the whole 2" screen with the rings sort of visible running around the edge. But hey this is the best I have seen yet! I noticed the 47 ohm resistor in the vertical section is black but it measured ok so I left it. I pulled the vertical output assembly out of my other Starcastle and bam! I got a full screen, however it was VERY dim! I then fired up the Picture tube tester and found the emissions test to be at 0, the needle only twitched! I performed a clean operation a few times and got the emissions up to "OK" but weak, I did not want to hit Rejuvinate as it may destroy the tube. I hooked everything up and the screen was excellent! In fixing this game, I can troubleshoot the first game with no vertical deflection as I now know the problem is before the final output transistors. The second game has a much odder video problem, I'll have to think about that one a bit more.

_______________________________________________________________
6/18/00 Repairing a Devastators PCB
This was a simple one. It came up and said "bad ROM K11". I burned a new ROM from a good ROM image and this fixed the game.

_______________________________________________________________
6/15/00 Repairing an Aero Fighters PCB
The game seemed to play but the video was completly scrambled and out of sync. If I flexed the PCB the video would change but not improve. I then looked around the surface mount chip at U43 and found loose pins around all four corners of the chip. Resoldering these pins corrected the problem.

_______________________________________________________________
5/21/00 Repairing a JR Pacman PCB
This board had a display that looked like there was no horizontal or vertical sync. I looked at the frequencies on the various counters (74161/163s) and nothing made sense, the freq counter simply displayed random frequencies. I then went to the clock circuit and it was doing the same thing. I found that the scope couldn't get a lock on the waveform yet I got a good signal, just not stable. Looking closer at the board, I found a leg of cap C4 (1000pf) in the oscillator circuit was broken from the PCB. Resoldering this fixed the board to playing, but there was a graphics glitch. Some of the various characters had snow and wrong colored lines running through them. I swapped the IC at 6C which was an MB8128-15 with a 6116 RAM (same chip, different part number) and this fixed the board.

_______________________________________________________________
4/24/00 Repairing a Space Invaders
All I could get out of this game is garbage on screen. The first thing I checked with the scope is the reset line and found that it was being held in reset. Tracing around the ICs I found that the power on reset was held low. I replaced the IC at F3 (74LS04) but now I was getting intermittent resets. I checked the power on reset counters and buffer chips but did not find any reason for this activity. I gave up trying to figure why so I did a modification to the reset circuit (I eliminated the POR from the power supply). To do this I lifted pin 10 of IC F3 from the socket. Then I soldered a 1.5K pullup resistor (to +5V pin 14) onto pin 10 of the socket. This fixed the random resets, but I could not get any coinup at all. I traced the inputs to IC C2, a 74LS14, and replaced it. This fixed that problem. Now I found that the shoot sound was missing. Replacing the LM3900 at K4 fixed this. So all of the problems resided on the sound PCB this time. Thanks to Todd Moon for allowing me to fix his PCB. Update, so far we cannot get this boardset to work in his machine, I'll keep you posted.

_______________________________________________________________
1/20/00 Repairing a Defender
This game gave a consistent RAM error 12. This indicates a bad RAM chip at location 4M (bank 1, chip 2). Here is a list of the Defender RAM error codes. Replacing this chip fixed the RAM error. Thanks to Layne Goetzinger for sending feedback on the repair.

_______________________________________________________________
12/28/99 Repairing a Defender
What would happen to this game was it would rarely come up properly. It would give a variety of RAM errors or if it was running, it would lock up or reset. Re-seating the RAM and swapping out the RAM did not fix the problem. Looking at the +5V line was getting 4.6V at the RAM chips, too low! Running heavier guage wire directly to the RAM/CPU board brought the voltage up and the game now works. Thanks to Gary Kueck for the fix.

_______________________________________________________________
11/24/99 Repairing a Crazy Kong Part II board
This is a Donkey Kong clone from Falcon. This board worked for the most part except kong had scrambled graphics along with some other graphics glitches. Poking around and pressing on the ROMs at locations 11A and 11C, I found a broken pin on ROM 11C but the graphics was still jittery. So I replaced the socket and that fixed the problem.

_______________________________________________________________
10/4/99 Repairing a DR Mario board
There was no picture at all on this board. Poking around with the scope I found that a bunch of the data lines on the processor were floating. Pushing on the satellite board affected the lines. I then looked with the scope on the RGB resistors (R47, R46, and R89) and sync (R30) and found signals! I proceeded to hook the game up to a monitor and found it working! cleaning the contacts on the satellite board fixed the problem. It was in free play so I switched SW2 # 7 to off, now it was in normal operation.

_______________________________________________________________
9/27/99 Repairing and Asteroids Deluxe
During self test the game gave a high then a low tone only. This indicates a bad RAM at M1. I replaced the 2114 which fixed it. Here is the error code list.

_______________________________________________________________
8/16/99 Repairing a Sega Astro Blaster
This is a cabaret machine that plays just like Atari 2600 Megamania by Activision! This game has voice but there was no other sound output. I simply replaced the sound board which fixed it. I will fix the bad board as I have time. Since there is no sounds I guess the final Op-amp section on the board is bad.

_______________________________________________________________
6/8/99 Repairing a Continental Circuit (Circus)3D Racing Game
The main problem is there was no horizontal deflection, just a thin vertical line. The game coined up ok. The monitor is a Wells Gardner 19k7901 monitor. Looking at the PCB I found that some resistors got hot and the horizontal coil had charred contacts. Re-soldering the coil and replacing R95 (2.2K 1W resistor) across the coil fixed the monitor. A cold solder joint on the horizontal coil had caused the damage (burning open the resistor). I am still working on the steering wheel, it needs a new centering spring and the rubber end stop needs to be re-mounted. The game is at least playable now. Update, I added the centering spring and rubber end stop but this was more trouble than it was worth. The game would "drift" out of steering alignment, you would be at the end of the rubber end stop just to go straight after playing awile. So I removed the spring and stopper and it plays fine.

_______________________________________________________________
6/1/99 Repairing another Tempest PCB
This is a board that had two problems. The first was that it was stone dead, no output. This was an easy fix, I simply re-seated the ROMs. Now the PCB had random vector lines running all over the place during attract and game play, however the Tempest logo and High scores looked ok. I had a spare AUX PCB so I swapped out the 40 pin chips (math chips). It turned out to be a curled under pin on one of the chips.

_______________________________________________________________
5/11/99 Repairing a Tempest PCB part II
Well I found out that if I don't hook +5V to the AUX PCB, the game will act as if it's in free play no matter what the DIP switches are set to. Here is a picture of the scope output of the fully working Tempest PCB.

_______________________________________________________________
5/6/99 Repairing a Tempest PCB
This board had vertical, but no horizontal deflection. Since I do not have a vector monitor or the Atari power supply, I had to come up with a way to use my scope and a computer supply. To do this there are two things, the power for the vector outputs and the reset line using +10V for the POR circuit. The first thing you must do is to desolder/clip the outputs (pin 3) of the 7815 (VR3) and 7915 (VR1) regulators away from the PCB. Then you can hook +12V and -12V to the +15V and -15V testpoints, also don't forget +5V and GND. Next you will have to ground pin 2 of IC E6 (hex inverter). This is a simple way of getting the board out of permanent Power On Reset, you don't need to fiddle with the +10V. Now I simply hooked my Channel 1 to Y OUT and my channel 2 to X OUT testpoints (set the scope to XY). Now to fix the board, I found that I had a good X input to IC C12 (SG1495), but no output. Replacing this IC fixed the problem. Notice the retrace lines as there is no Z shutoff to kill the beam.

_______________________________________________________________
3/31/99 Repairing a Centipede PCB
This board would not come up at all, there was no activity on the 6502. I found there was no clock at all (12MHz or 6MHz). I replaced the 12.096 MHz crystal and I got a working Centipede, sort of. Along the right side from the middle to the bottom of the screen there were letters running down, some were cycling the alphabet. Well there are eight 2101 ICs to try so I decided to test for the bad chip using the "piggyback method". What I mean is you get a good RAM chip and press fit it over each chip, If the video improves or changes, that chip is your best bet. I found that piggybacking IC K7 nearly fixed the display, a major improvement. Replacing this IC fixed the game.

_______________________________________________________________
3/29/99 Repairing a JR Pacman PCB
This is a board my friend bought off of Ebay. It had some smoke damage but it was cleaned somewhat. This board is similar to Pacman electronically, not physically. There was no video output at all and the signals around the Z80 had no activity except the clock in on pin 6. Poking around with the scope I found that the counters at 2R and 2S were dead even though there was a 6MHz pulse going into them. I found that replacing 2R got all of the counters running and the Z80 was cycling now, but still no video. Reseating the ROMs got the game to come up but we have no sound. Looking at the LM1877 we found that it had an internal short from +12V to GND. This explains the smoke from the chip when +12V was applied ;). This board is fixed save for the now missing LM1877. Update, replacing the LM1877 restored the sound.

_______________________________________________________________
3/17/99 Repairing a Galaxians PCB
This board did not display anything at all. The Z80 had no signals other than the clock. I first looked at the reset circuit and found it stuck in reset. IC 6E (7404) had pin 10 and 11 stuck high. IC 6N had no valid outputs. I replaced these and now had garbage that would show, then fade away. I found that IC 5C (7474) had no output where a 3MHz pulse should be. Replacing this restored the puse but it didn't change anything. Poking around in the video RAM area I found that pins 2,3, and 4 of IC 5J (74245)had invalid TTL outputs. Replacing this gave me steady garbage now. Removing the program code satellite board (at row 7) did not change the garbage at all so I guessed that no code from that board was running. Yup, all of the ROMs were bad. I now had a working game, sort of. There was no green, the "GAME OVER" was not visible, the "WE ARE THE GALAXIANS" was blue not red, and parts of the objects were missing. I troubleshot this the hard way, I found that there was no signal at VID1. Tracing back I found no RAW1 at 3L, no LSD1 and MSD1 either. It turned out to be a bad character ROM at 1H/J. The game is now working, I have no idea how it got into such a bad state (lightning?).

_______________________________________________________________
1/17/99 Repairing Asteroids
This game would only display four rocks frozen around the edge of the screen and would not coin up or free play. There was also a row of "C"s across the bottom. The game board worked fine in test mode, no errors at all. Looking around the board for missing signals, I found that IC K5 (74LS191) had floating outputs on pins 2 and 3. This chip is the up/down counter for the stack register in the vector generator program counter circuit. Replacing this chip fixed the game.

_______________________________________________________________
12/26/98 Repairing Uniwar S
This game died again. The thermal problem progressed (see below). The game would give scrambled video, then it eventually died with a black screen. Using the O-scope I found a 74LS10 at location 5D to have no output on pin 8. This was signal *LD on the 3 input NAND gate. I now had garbage on the screen again. looking around I noticed that the ROMs on the satellite board were not being selected. Removing the satellite board gave the same garbage so I figured the ROMs were bad or something. Nope, It was a bad 7442 IC on the satellite board.

_______________________________________________________________
12/26/98 Repairing a Galaxians board
This board would only show garbage. The ROMs were very corroded, so bad that when I tried to remove them the pins snapped off. So I replaced the sockets (broken pins in them) and pulled a set off the web to burn a new set. The new set got rid of the garbage but now I was getting a "BAD RAM #3 error. Replacing the 2101 RAM IC at location 5F fixed galaxians.

_______________________________________________________________
11/5/98 Repairing Cheyenne by Exidy
There were multiple problems on this machine. One, it didn't have a power supply. I installed a PC power supply to replace the missing supply. I found that there was no video, just a white screen. I took the board to my bench and found that there was no CPU clock on either 6809 CPU. I checked the crystal and it was oscillating. Poking around and tracing the clocks around using the schematics I found on SPIES.COM I found a loss of signal at IC 6J. This was a 74s161 4 bit counter. Replacing this restored the games logic. I could see action on the o-scope now. However I still had no video. I then remembered to hook up the 12VDC which fixed that problem. There is a bank of transistors driving the video circuit that requires 12V. Note there is a 7812 voltage regulator that I had to jump out to use the computer power supply. After fixing those problems I found that there was no sound on the left speaker. I found a bad TIP120 transistor on the sound PCB.

_______________________________________________________________
10/26/98 Fixing Eagle by Centuri
The symptoms with this game was poor vertical deflection. Since the monitor is a vertical mounted monitor, the left and right sides were shrunk and folded over. Vertical hold also was drifting around. The monitor board is a Toei CM-A20HC (the PCB had CM-B14 silkscreened on it). I first replaced C83 which is a 10uf/350V cap which filtered out one of the DC power lines from the flyback. This fixed the vertical hold problem but the vertical was still shrunk. I then replaced C62 which is a 220uf/16v cap. This capacitor is just along the vertical amplifier path near the vertical width pot. This fixed the shrunken vertical picture.

_______________________________________________________________
8/12/98 Fixing Another Asteroids Deluxe
This game would run a few minutes and then the display would bloom and fade out. Replacing the HV diode fixed the problem. The diode is no longer being produced and one was found at a Fry's electronics (the last one they had). Thanks to Richard Pike for this repair information.

_______________________________________________________________
8/4/98 Fixing Planet Wars (Galaxians clone/Uniwar s)
The game would simply cycle through the diagnostics over and over. Looking at the screen you could see a brief message saying "BAD RAM 3", this was hard to see, you would miss it if you blinked. There is no reference on the board to RAM 3, so I started swapping RAM chips around. There were two 2101 ICs at 4F and 5F next to each other. I swapped them and the diagnostics changed from red letters to green letters and still said bad RAM 3, I now had an idea that it was one of these. I found a 5101 IC on another game board and swapped out IC 4F. I'm guessing this is a good substitute for a 2101. This got the game going sort of. I now seem to have a thermal problem but I believe it's a weak power supply (it wasn't).

_______________________________________________________________
8/2/98 Fixing Asteroids Deluxe
This game had a video problem where the video would bloom and fade away. Banging on the HV box would sometimes bring it back. Resoldering all connections on both the deflection board and the HV box and cleaning contacts fixed the problem. The video board had many cold/bad solder joints as well as corrosion on the transistor sockets.

_______________________________________________________________
7/7/98 Fixing Shinobi (part 1)(Sega)
Just about everything was bad on this machine. Bad monitor, bad power supply, and bad main logic board. I fixed the first two and will have the MLB repaired as soon as the part comes in (the audio amp IC MB3733). The first problem was that the power supply was dead. I found that the fuse was blown so I took a few measurements around the bridge rectifier. The rectifier was shorted on one side so I replaced it and the fuse. This fixed the power supply. I also had a second power supply from a box of junk from the auction and fixed it by replacing the two filter caps for the 160VDC line (just after the bridge rectifier). This second supply would blink the +5V LED on and off, it would not put out +5V (noisy). Looking at one of the caps, I noticed that it had swelled and leaked. Replacing the cap fixed the power supply. The monitor board was fixed by installing a cap kit and fixing a broken wire along the 135V line (see Wells Gardner below). The MLB has no sound output but the input is there (on the o-scope). Replacing the audio amp will fix this I believe as I have all required voltages and sound going in.

_______________________________________________________________
6/3/98 Fixing Astro Fighter (part 1) (Sega/Gremlin)
This game gave a black screen or a white screen randomly, with no actual picture. I measured the voltages and found +5v varying between 1v and 3v. Disconnecting the game boards allowed +5v to come up. Replacing the 723 voltage regulator caused the power supply to put +12v on the +5v line (glad I was using a dummy load!) I then replaced the 2N3055 power transistor which fixed the power supply. However the game does not come up properly. I get a sound effect, then the starfield and part of the player/score box at the top, then it locks up. Oh well, I do not have an EPROM programmer that does 2708 ROMs to try and fix it. I will try a few more things before I trash the game.

_______________________________________________________________
3/9/98 Fixing Bosconian
I Bought a box of game boards from Great Western Trading. I bought 35lbs worth which gave me 12 game boards (great deal!). So far I have repaired 7 boards. The first is Bosconian. This board gave an error of "BAD RAM 4L", so I started swapping RAM around. There are three 2016 RAM chips on the board and none of them are at location 4L. Swapping one RAM from the video board to the CPU board caused a different RAM error. (1E which still doesn't point to a correct location). I then replaced the RAM chip and got the game to boot. Also there is severe corrosion on the other socketed chips. I tried to remove one of them and a pin fell off of it. I had to solder a new pin on it. So cleaning the pins is not an option currently. Oh well. I will write more info as I have time so check back again.

_______________________________________________________________
1/20/98 Fixing Guerilla War (SNK)
Well I gave up for now on the Wells Gardner video board so I substitued an Electrohome G07 board. I had to replace the yoke, re-adjust the purity magnets, and adjust the R, G, and B drives to get a good picture. So I now know that a G07 will drive a monitor from a WG with a little tweeking.

_______________________________________________________________
1/9/98 Gyruss Repair
Gyruss developed a problem with the generation of the sprites. It looked like solid horizontal lines through them. I found out that swapping the 086 chips at locations 3G and 7F caused a color palette problem, but the sprites looked ok. These chips are sanded off so I'm not sure what they are (Mux chips?). The next step was to make the game work with what I had. It turns out that the chip at location 3G uses only half of the chip so I created a circuit board to route the working half of the bad chip into the palette circuit. This fixed it!

_______________________________________________________________
12/19/97 Tempest Repair
Shane Moura writes: As far as using my experience of fixing Tempest on your site, I think the real credit goes to Nick Sayer and his Tempest site. Although he didn't have any mathbox info, Gregg Woodcock wrote an article that gave me the clue I needed.

Suffice it to say: You can suspect a mathbox failure if the self test comes up with an M above the grid. Also, the mathbox is only used in the perspective calculations of the game. If the high score table and the title screen displays correctly, but the game attract mode and the actual game itself is rendering wrong (ie: jumping around all over the place) then the mathbox is probably screwed.

You replace it with AMD 1901 ALU's, it takes four, and they cost about $10.00 apiece reburbished, $15.00 apiece new. (not $1.50 like it says on the Tempest site.) They are socketed, so...

Disconnect the power cables from the double board assembly. Remove and separate the boards. Remove the offending IC's on the smaller board. (At locations E2, F/H2, J2 and K/L2). Pop in the new 1901's, and BLAM... You're rollin!

_______________________________________________________________
12/18/97 Repaired another Defender.
Re-seating the various ribbon cables between the boards fixed this one.

_______________________________________________________________
12/18/97 Fixing Space Invaders.
It seems that the crystal fixed the unit. It is a deluxe MLB with an original SI ROM set. The crystal was a 19.96MHz crystal and we are now using a 20MHz crystal (you have to adjust the sync on the monitor a little). A deluxe ROM set is going to be installed as soon as they are burned. The MLB will support 2716 ROMs with a few jumper changes.

_______________________________________________________________
12/17/97 Fixing Space Invaders.
We are just starting on this one. So far found that the 19MHz crystal is bad. The game had an assortment of boards in it, both Invaders and Deluxe boards floating around in it.

_______________________________________________________________
12/16/97 Fixing Eswat.
There was no horizontal deflection, just a thin vertical line only. There was also no sound. I found that the horizontal width coil was broken from the video monitor board (a G07). Resoldering it fixed it. The sound problem was because of the +12V was missing. I took apart the power supply and found the noise coil broken from the PCB (two broken coils?? dropped unit?).

_______________________________________________________________
11/09/97 Repairing Gorf.
Well I finally fixed the problems with Gorf. The first problem was the video. I was missing blue from the video. I found that the problem was on a small PCB mounted in the case. This board has one IC, a handfull of resistors, and two delay lines. The video from the main PCB is sent to this board and the output is sent to the Electrohome G07 monitor. I found that the delay line for the blue was open. I did not have a replacement so I jumped it out. The video looks pretty good now even without the delay line.
The second problem was with the lights that tell what rank you are up to (cadet, captain, etc.). One light refused to work. Replacing a TIP 110 transistor on the light board fixed the problem. I substituted a TIP120 for the TIP110. I had tried some other transistors in this, but they refused to work (need a high power switching transistor for this). The circuit is simple. The transistor simply connects ground to the particular light allowing the 10V to pass through.

_______________________________________________________________
11/05/97 Fixing Trivia 2.
I now have both Trivia Quiz games working. This one had no video also, but it was the MLB causing it. The problem was a dirty socket for the TI video chip. Re-seating the chip fixed the game.

_______________________________________________________________
11/05/97 Fixing Defender.
I replaced the monitor with a Wells gardner from Street fighter II. The original monitor was a G07 with a broken tube. The monitor simply plugged right in and ran. Before I did this I was using a Hanterex from the broken trivia game. Warning if you dont use isolation on the monitor you will smoke a resistor on the Defender PCB (I did this)!

_______________________________________________________________

If anybody has any other tidbits to add, Email me and I will add them here.