So, I'm just getting back into games after a long while (like, RGVAC was cooler than a web forum long).
Anywho, I decided to dig into this stack of dead boards I've been carrying around with me and I figured I'd pick some expert brain about my problem children.
These are all on a test bench with a switcher and a JROK RGB to component board except for the vectors, which are in a space duel cabinet. I've got a USBee locic analyzer for poking around.
Set 1: Robotron:
This boardset is dead, no video, no 7 segment LED, watchdogging like craziness. I put new sockets in for the decoder ROM 6 (corrosion), the CPU (corrosion), and replaced the RAM with 4164's. I can hold reset and watch the CPU write to the reset address, see ROM 12 get selected and write data back to the CPU. Taking a look at 5G to see what's going on in the watchdog ckt, I see that pin 3 is held low. Tracing that all the way back to 4G, I see why. A10 and A11 are in lock step, so that pin never gets decoded. Frustrated, I'm thinking, Why no video? I can trace that all the way back to a Screen Control signal that's doing nothing. That ends for me at the 6821.
I need to walk away from this set for a while, or at least until I can pick up another set to compare against. But if anyone has any suggestions..
Set 2: Centipede:
Havent touched this in a while, but it's next when I get a new edge connector. Plays blind, but I get nothing but a yellow screen. Any preliminary suggestions?
Set 3: Tempest:
I got this set when I was young and a hot shit, thinking I could fix stuff. I replaced all of the sockets and ram (didn't use sockets!) and apparently made a mess of the whole shebang. This set should probably be scrapped, but it's one if my firsts, and now its a challenge. I went around and pulled all of the sloppy sockets and poorly installed vector RAM out. I then went ahead and fixed all of the traces I ruined way back. Checked for shorts and continuity on the soldered in program RAM since I don't want to mess with that yet, I slapped new vector RAM and sockets in, and put one program ROM socket back in with a 27256. This is the way it sits and is, of course, watchdogging. Now I remember back when I was fixing a couple of Asteroids boards you could split the VSM from the program side to at least get that working first. Is this the case with Tempest? I'd love to be able to just the the program to start with no vectors or aux board and ease myself into this one.
Anywho, I decided to dig into this stack of dead boards I've been carrying around with me and I figured I'd pick some expert brain about my problem children.
These are all on a test bench with a switcher and a JROK RGB to component board except for the vectors, which are in a space duel cabinet. I've got a USBee locic analyzer for poking around.
Set 1: Robotron:
This boardset is dead, no video, no 7 segment LED, watchdogging like craziness. I put new sockets in for the decoder ROM 6 (corrosion), the CPU (corrosion), and replaced the RAM with 4164's. I can hold reset and watch the CPU write to the reset address, see ROM 12 get selected and write data back to the CPU. Taking a look at 5G to see what's going on in the watchdog ckt, I see that pin 3 is held low. Tracing that all the way back to 4G, I see why. A10 and A11 are in lock step, so that pin never gets decoded. Frustrated, I'm thinking, Why no video? I can trace that all the way back to a Screen Control signal that's doing nothing. That ends for me at the 6821.
I need to walk away from this set for a while, or at least until I can pick up another set to compare against. But if anyone has any suggestions..
Set 2: Centipede:
Havent touched this in a while, but it's next when I get a new edge connector. Plays blind, but I get nothing but a yellow screen. Any preliminary suggestions?
Set 3: Tempest:
I got this set when I was young and a hot shit, thinking I could fix stuff. I replaced all of the sockets and ram (didn't use sockets!) and apparently made a mess of the whole shebang. This set should probably be scrapped, but it's one if my firsts, and now its a challenge. I went around and pulled all of the sloppy sockets and poorly installed vector RAM out. I then went ahead and fixed all of the traces I ruined way back. Checked for shorts and continuity on the soldered in program RAM since I don't want to mess with that yet, I slapped new vector RAM and sockets in, and put one program ROM socket back in with a 27256. This is the way it sits and is, of course, watchdogging. Now I remember back when I was fixing a couple of Asteroids boards you could split the VSM from the program side to at least get that working first. Is this the case with Tempest? I'd love to be able to just the the program to start with no vectors or aux board and ease myself into this one.
Last edited by a moderator:


