Asteroids Troubleshooting - Need help

orion3311

Well-known member

Donor 2013
Joined
Nov 25, 2008
Messages
5,217
Reaction score
30
Location
Blue Bell, Pennsylvania
Figured I'd make a new thread - been trying to learn how to troubleshoot these things and I have Ray (Mundoman's) Asteroids board here. I went away last week and before I left, it would produce the test screen (with watchdog disabled) but in both normal and self-test modes the processor is watchdogging. In game mode, it would start to draw something and watchdog...no actual game or anything beyond what looks like maybe a 'P'.

Now that I tried it again, self test doesn't seem to work either (not sure what happened??), however if I push the dmago signal by writing to 0x3000 on the Fluke, it draws the self test screen with a brighter line going top-right to bottom-left behind it and flickering.

Connected the Fluke 9010a and all ram tests pass, as well as all the roms and proms. I also used the fluke to verify all the switch and dip switch inputs and they read perfectly.

I originally suspected the program counter at J4, and tried to piggyback one but that didn't seem to do anything.

Where to go from here??
 
Wow Jeff, can't help you on that one- you've gone way beyond my meager skills!

Up until this post, I always thought Asteroids boards were low-maintenance and pretty efficient. This is a real eye opener.
 
Last edited:
Thats been my experience as well, but like anything else they *do* fail and in this case it was already repaired by a "board repair company" and failed again.

I figured I'd try to do this as a favor to Ray while learning the ins and outs of it but man this board is a LOT more complicated than the simple ones I'm used to.
 
FYI - at this point...

Via the Fluke 9010a...
-----------------------
Program and Vector Rams test good (Short and long tests)
Roms test good (also verified via ROM burner but still suspect- see below)
Dip switch, coin door & control panel switch registers all read properly
Sounds all function properly (writing to various sound addresses produce proper sounds)
Due to all above tests, I'm assuming all address and data buffers working properly

Lifting DMAGO did not stop watchdog, which leads me to believe its still a logic problem.

ONLY odd thing I see, is that reading the self test switch at 0x2007 produces FF whether the switch is on or off, YET the board recognizes the switch properly somehow. I'm suspicious of this but not sold thats part of the problem.

Only other thing I noticed, is that despite all the roms verifying, the "repairer" used 4 different roms!! There's literally a 2716, a 2732 (double imaged), an 8532, and a mask rom! WTF! I'm going to see if I can find some 2716's and make a nice clean set to try before moving on to other things. While these may all work (or did work) I can definitely see timing issues with 4 different brands/types of roms.

Last but not least - does anyone have notes or schematics of the "unpublished" portions of the board, like the switch matrix at L10?? Seems they left a lot out of the schematics and thats why I'm struggling to figure this out.
 
I believe all the circuitry is documented in the schematics - what section are you missing ?
 
I don't see anything regarding how the switches interface to the logic portion, like the 74LS251 @ L10?

I burned a new set of roms - no go.

I did discover one thing...I put the CPU back in and the self test makes a quick low beep (although still watchdogging). Turns out the Fluke only has a 6502, not a 6502A.
 
Ok, apparently I'm just a moron...I thought I downloaded all the sheets, all 180 times I apparently neglected to get the last sheet. Son of a....
 
Now that I have the whole picture - turns out that chip @ L10 is bad so far. I moved the other one at L11 over to L10 and now reading the self test @ 0x2007. Its also now writing dmago again but still not working properly. Not sure if its because the chip @ L11 is missing or a problem with the state machine yet.
 
Sooo...I randomly checked the pins on the 3 program counter chips, and found pin 12 dead. Traced that back to pin 3 @ C6. Pins 1 and 2 were pulsing ok, which made C6 (74LS00) suspect.

Fired up ye olde Logic Comparator and pin 3 is blinkin at me. Piggyback a new 74LS00 on top of C6, and I gots me some Asteroids on my O-Scope :)

Can't play or overall test it yet as L11 is still missing (gotta find or get a 251 to replace it), but it appears to be working ok, and its running on the Fluke so I guess its ok.
---------------------------------------
So...symptoms were originally Asteroids watchdogging in both game and test modes. It was drawing SOMETHING on the screen in both modes, but after a while even that stopped.

The switch input mux @ L10 was dead, showing both HALT and TEST being constantly active. I moved L11 over to L10 and both signals started working properly. L11 isn't required for it to run because it handles the coin door and control panel inputs. It won't play but it'll show attract screen at least.

Now its drawing what looks like the test pattern with a diagonal line through it (in game mode). Another repair log mentioned a simliar symptom and to look at the program counter (74LS670) at J4, so in checking that I found pin 12 dead.

Pin 12 comes from pin 3 @ C6, which is a 74LS00. Pins 1 and 2 were pulsing and I verified Pin 3 was malfunctioning with a logic comparator. Piggybacked (for now) a new LS00 on C6 and board came to life.

Goodnight!
 
This is way beyond what I understand, but I think it means I might not have to send out the board for repair. If that is right thank you VERY much!!

Sooo...I randomly checked the pins on the 3 program counter chips, and found pin 12 dead. Traced that back to pin 3 @ C6. Pins 1 and 2 were pulsing ok, which made C6 (74LS00) suspect.

Fired up ye olde Logic Comparator and pin 3 is blinkin at me. Piggyback a new 74LS00 on top of C6, and I gots me some Asteroids on my O-Scope :)

Can't play or overall test it yet as L11 is still missing (gotta find or get a 251 to replace it), but it appears to be working ok, and its running on the Fluke so I guess its ok.
---------------------------------------
So...symptoms were originally Asteroids watchdogging in both game and test modes. It was drawing SOMETHING on the screen in both modes, but after a while even that stopped.

The switch input mux @ L10 was dead, showing both HALT and TEST being constantly active. I moved L11 over to L10 and both signals started working properly. L11 isn't required for it to run because it handles the coin door and control panel inputs. It won't play but it'll show attract screen at least.

Now its drawing what looks like the test pattern with a diagonal line through it (in game mode). Another repair log mentioned a simliar symptom and to look at the program counter (74LS670) at J4, so in checking that I found pin 12 dead.

Pin 12 comes from pin 3 @ C6, which is a 74LS00. Pins 1 and 2 were pulsing and I verified Pin 3 was malfunctioning with a logic comparator. Piggybacked (for now) a new LS00 on C6 and board came to life.

Goodnight!
 
This is way beyond what I understand, but I think it means I might not have to send out the board for repair. If that is right thank you VERY much!!

Thats what it means - I just need to find one chip (might be able to pull it from a junk board) and it should be good to go if it works ok.
 
Thats what it means - I just need to find one chip (might be able to pull it from a junk board) and it should be good to go if it works ok.

You forgot to mention Jeff that that one chip actually has 10K gold inlays, and can only be found at a jeweler's store because they are so rare.. :D
 
Turns out both 251s were bad, and I found some on a scrap board. Its kinda difficult to play Asteroids on an o-scope with nothing but a clip lead for a control panel, but I was able to almost clear the entire first board doing nothing but shooting to the right LOL.

Ray if you're going to York this weekend let me know I'll bring it with me, OR I'll stop out next week and we can verify the voltages again.

(Believe this board was killed due to the power supply being set too high by the former owner).
 
Back
Top Bottom