Asteroids weirdness - thoughts?

modessitt

Volunteer: Encyclopedia Submission Moderator
Joined
Mar 22, 2007
Messages
33,414
Reaction score
325
Location
Austin, Texas
So I've had this Asteroids in the shop for a few weeks. Originally it had some problems that I repaired:

http://forums.arcade-museum.com/showthread.php?t=116746

The game has been running fine for the past two weeks. Friday, we played a bunch of games on it with no problems.

Monday, I delivered it to a customer and it came up fine. We played a few games on it and everything was ok. As I'm standing there talking to her, the screen goes blank and the start buttons are flashing rapidly. I pull the game back out and open the back door. I reseat the connector, push on some chips, and turn it back on, and the game is working again. I put the back door on and put it back against the wall and the screen goes out and the buttons are blinking again.

I pull it back out and check voltages. Everything is fine. While I'm sitting there thinking, the screen comes back on and everything is fine. After about a minute the screen goes out and the buttons are flashing fast again.

Now, it won't come back up. Spotkiller is on, LEDs flash very fast, and nothing. +5 is perfect. Flipping to test mode gives no beeps. Asteroids Encyclopedia mentions K5 can cause this. Replaced it and it didn't help.

I tried the board in another working game, and I get the same result, so I know it's a board issue. Also, I swapped all the socketed chips from the good board to this board and it didn't help, and the good board still works with these ROMs and 6502.

Anyone got an idea??? :confused:
 
ast

id lightly sand every socketed chip leg front and back with the smallest grit metal oxide sand paper out there
 
id lightly sand every socketed chip leg front and back with the smallest grit metal oxide sand paper out there

Did that. No help. Changed out chip sockets. No help. Swapped chips with known-good set. No change...
 
Well the first thing I would try is checking the NMI circuit output on a scope. Make sure you are getting pulses.

After that, you'll just have to isolate the DMAGO line, and see what half the problem is coming from.
 
i think its a dried out electrolytic capacitor. id install new ones and see. if same problem after, id turn it off and feel if any transistors are getting hot and replace.
 
Well, I don't have a scope, but I have a logic probe. While checking something mentioned in the Asteroids Encyclopedia, I find that pin 5 of the LS161 at D7 is low when it should be pulsing. It is fed by pin 2 of the LS42N at E7 which is also low when it should be pulsing. Not sure if this is the problem, or if it is even farther upstream.

Now, I've got the schematics, but I'm not too conversant on signals and when I should expect a pulse and when I shouldn't.

Can anyone clarify?
 
Well, I don't have a scope, but I have a logic probe. While checking something mentioned in the Asteroids Encyclopedia, I find that pin 5 of the LS161 at D7 is low when it should be pulsing. It is fed by pin 2 of the LS42N at E7 which is also low when it should be pulsing. Not sure if this is the problem, or if it is even farther upstream.

Now, I've got the schematics, but I'm not too conversant on signals and when I should expect a pulse and when I shouldn't.

Can anyone clarify?

Keep going backwards... but I think you've got it. A 74LS42's purpose is to take a 4-bit (binary-coded decimal) input and turn it into a single pin's (as in "one of ten output pins") output. I don't see why anything should be special (i.e. always low) about one of its output pins.

If you've got pulsing on all four pins (12/13/14/15) of E7, it's probably E7.

else-if you've got static input on those four pins (12-15 of E7) that corresponds to an output of "1" (pin 2 of E7), you might have to look further upstream. (The Scale0/1/2/3 inputs to M7 should always be pulsing as well as the 1/2/3/4Y outputs of F6, as the vector generator gets instructions to draw big/medium/little rocks, or letters/numbers.)
 
Thanks for that. I'm going to print that out and take it with me to the board.

As for my prior post, the Encyclopedia says that pin 5 of D7 should be pulsing (it's low), but looking at the drawing package shows that Pin 5 is ground, so it should be low. Still, I'll check those input to E7 for pulses and see what's going on...
 
It's just another example of machines "paying you back" for selling them to someone else. They become comfortable when they live with a person who loves them, nurtures them, makes them all sparkly clean with magic eraser... they get used to that life.

But when you rip them from that warm cuddly womb, break yoself. They bite back. Believe me when I say that there's some kind of hex there. Both my Defender and Robotron did it to me!
 
Well, this is a work game. Came in broken from one person and is going out fixed to another person.


Found no pulses on any of the SCALE inputs to M7, and none of the 1/2/3/4Y outputs of F6. Further testing revealed no pulses on any of the Timer 0/1/2/3 inputs to this circuitry, so I have no pulse on the CK input to pin 9 of M7. I'm guessing this is a problem.

I'm about to check the LS273 at F7 where the TIMER0/1/2/3 comes from. I have verified that my crystal is pulsing, and I have a pulse on the -RESET line where I've checked for it. Then I'll check the ROM at N/P3 for outputs...
 
This shit is just too damn confusing.

I have no pulsing outputs of the LS273 at F7. All outputs (TIMER 0/1/2/3, DVY8/9/10/11) are either high or low. The DDMA inputs from the ROM are pulsing away. I have the pulsing signal at pin 1 and good +5, but no pulsing on the -LATCH 1 input.

Trace that back to the STATE MACHINE circuit and the LS42 decoder at E8. No pulses on input pins 12/13/14/15. I follow those back to C8 (034602) and find the inputs to that are fed by the TIMER 0/1/2/3 signals, which I already know aren't pulsing.

What did I miss? Is there another signal to one of these chips I forgot to check and could be causing my problem?
 
This shit is just too damn confusing.

I have no pulsing outputs of the LS273 at F7. All outputs (TIMER 0/1/2/3, DVY8/9/10/11) are either high or low. The DDMA inputs from the ROM are pulsing away. I have the pulsing signal at pin 1 and good +5, but no pulsing on the -LATCH 1 input.

Trace that back to the STATE MACHINE circuit and the LS42 decoder at E8. No pulses on input pins 12/13/14/15. I follow those back to C8 (034602) and find the inputs to that are fed by the TIMER 0/1/2/3 signals, which I already know aren't pulsing.

What did I miss? Is there another signal to one of these chips I forgot to check and could be causing my problem?

So far, I don't think you're missing anything.

PROM at C8: Yup, if the timer inputs are flat, you're not going to get anything interesting with the AND gate at E6. (Hmm, what if the AND gate is hosed? Hard to tell in-circuit...) TIMER signals feed E6, which feed the PROM at C8, which feeds D8, an LS174, which feeds E8, an LS42 which is really faded on my PDF, which spits out more -LATCH outputs.) So far, everything's consistent.

So, why are the timer inputs F6 flat? Because the outputs of F7, which you've already checked, are flat. But the DDMA inputs are good. And the -LATCH 1 is flat. (Are the -LATCH0, 2, and 4 inputs to the other LS273s also flat?)

I'd look at that other LS42 near D8. (And check the 3-inputs to the 3-input NAND gate at A8.)

(If you've got one from another board, I'm assuming you've tried replacing the PROM at C8 with one from another board, as opposed to reseating/cleaning contacts on the original PROM? Heck, if you've got a breadboard, let's confirm that there's actually still some data in that PROM...)

One other possibility -- are we looking in the wrong place altogether? Is -HALT stuck on B8? Maybe the state machine's fine, it's just never waking up, because something else on the board is resetting (via watchdog) the CPU.
 
So, why are the timer inputs F6 flat? Because the outputs of F7, which you've already checked, are flat. But the DDMA inputs are good. And the -LATCH 1 is flat. (Are the -LATCH0, 2, and 4 inputs to the other LS273s also flat?)

yes, all -LATCH inputs are flat.

I'd look at that other LS42 near D8. (And check the 3-inputs to the 3-input NAND gate at A8.)

5 and 3 appear fine. 4 is flat.

(If you've got one from another board, I'm assuming you've tried replacing the PROM at C8 with one from another board, as opposed to reseating/cleaning contacts on the original PROM? Heck, if you've got a breadboard, let's confirm that there's actually still some data in that PROM...)

Swapped a known good ROM set and there was no change.

One other possibility -- are we looking in the wrong place altogether? Is -HALT stuck on B8? Maybe the state machine's fine, it's just never waking up, because something else on the board is resetting (via watchdog) the CPU.

Hmm. Let me go check the -HALT signal...
 
I've seen similar behavior when Big Blue was bad. Don't go ripping your hair out chasing logic circuits until you replace that.

It really doesn't matter what the 5v line reads, if Big Blue is failing you'll get so much noise that higher level logic analysis is a waste of time.

Start with the simple stuff. Good smooth power is job one.
 
Thanks for the suggestions, but as I've already tested the board in another working machine, I know for a fact that power isn't the problem.

And for all of this we've talked about, the machine is now working again.

I decided to go back and check over a few things I'd done previously. Once I had this problem, info I found online initially pointed me to K5 (LS191) which I replaced. I had put a socket in when I did so and got no change, so I assumed it not to be the problem.

Well, I went back and did some continuity checks between the chip legs and the pads underneath and found a couple were not making contact. Problem turned out to be internal to the socket, so I pulled it off, put in another, put the chip back in, verified I had good continuity thru it, put the board in, and the game works! Just played a few games on it (high score 36,000 or so, I need my own game so I can practice more) and it's doing fine. I put the board thru some stress testing to see if I could get it to fail again, and it won't, so I'm hoping I have this problem licked.

FYI - I swapped the old chip back in and got the same error, so it appears that the LS191 at K5 decided to wait until it was delivered to fail. At least it happened while I was there, and not on the weekend during a party or something.

Will run it thru tomorrow to see if any errors resurface, then try redlivery tomorrow afternoon if all goes well.

Thanks for all the help, guys...
 
Oh man I hate when something like that comes back to bite you. Had a Berzerk where a socket wasn't making good contact. Spent hours trying to figure out what was up, and when I went back over all the worked areas checking continuity I found the problem.. Glad you got it up and running..
 
Back
Top Bottom