Space Invaders Deluxe

Holy crap I just about fixed board #3 - only have one Deluxe board left. I say just about because it works fine for 10-15 minutes then locks up. I think it has a bad ram somewhere yet.

A flakey ROM or it's socket can also cause this.....processor socket, too. Also, boarderline power.

Edward
 
I reseated the daughter board twice - first time it ran about 10 minutes, after that I reseated again and it ran to the point where I couldn't take the test sounds any more :)

I also think all this testing has taken its toll on the goofy power plug, it seems i really need to crank up the juice to get 5v at the board now, so I'm going to start using a clip lead.
 
Sounds like you've been a bit busy lately. You'll have all kinds of spare boardsets here shortly.
I only have the one boardset right now to play with.
Seems like you were maybe looking into the all-in-one ROM idea on another thread.
I was kinda curious about the multi SI setups, but hadn't looked into it much.

I'm still waiting on parts....... maybe I'll go clean the machine or something. :)
(actually I've been busy working on a Galaxian cocktail project while I wait)
 
Yeah I picked up a few that I figured were basically parts, but once I got one going I figured I'd try the rest, and so far so good!

The daughterboards though are all missing various sounds, and I'm not looking forward to those. I ordered a crapload of 3900s so hopefully they're all easy fixes!
 
Well, I progressively made good progress by just replacing RAM.... one by one as the test ROM spit them out to me.
First was "G" as it showed up most often.
That brought a somewhat sign of life back to it.
But then it showed "E", which I had already suspected.
Then "D" ....
Then it would finally show the test ROM screen with sound tests and all that. But after a couple cycles it would get glitchy.
Finally it shows a nice clear "A" ram as being bad.

Bam !!
It's game on time......... so I thought.
After letting it run awhile I came back to it being almost totally dead again, scrambled screen.
Voltages still check fine, test ROM won't show squat, doesn't matter if there's any other ROM's in in or not, changed CPU's, nothing changes. Gonna let it sit off a bit.
 

Attachments

  • Test ROM Bad RAM A.jpg
    Test ROM Bad RAM A.jpg
    51.7 KB · Views: 15
  • Test ROM Good.jpg
    Test ROM Good.jpg
    56.9 KB · Views: 13
  • Game A.jpg
    Game A.jpg
    50.7 KB · Views: 13
  • Game B.jpg
    Game B.jpg
    47.3 KB · Views: 14
Oh, and mine did NOT like S2 tied to ground at all, wouldn't do anything in that configuration.
So I set it back to the 5v spot. It would run with it disconnected as well.
 

Attachments

  • SAM_1074.jpg
    SAM_1074.jpg
    96.4 KB · Views: 14
I did have one small trace I hose up pulling the last ram chip off, but I think I did an ok job repairing it.
 

Attachments

  • Trace Repair.jpg
    Trace Repair.jpg
    92.4 KB · Views: 20
I cant see the ram repair with my phone but go back and double check your soldering to make sure theres no bridges.

As far as it crapping out...try reseating thw daughterboard a couple times, mine was resetting a lot and reseating the daughterboard fixed it.

Also if it seems heat related, you can use a can of duster (air) turned upside down as freeze spray to see if a certain area affects it.

Ps only leave the test rom in there until it runs reliably.

Jeff
 
Last edited:
I just tried re-seating the daughter board, no change.
No solder bridges.
Pulled all the ROMs, same screen. Verified ROMs still good.

The only time the screen changes is if I pull the CPU.
There are two areas or "strips" that are kinda in motion...if I pull the CPU those two areas go blank, but the rest stays the same.
Gonna double check the 40 pin socket again.

Doesn't seem heat related as I left the machine off awhile and it's still the same garbage.

Need to verify how to test the reset line.
According to the power supply test page I was using http://www.elektronforge.com/TestMdwPS.htm there should be less than 0.5v... well, I have 0.01v, that's definitely less.
I don't have an oscilloscope to further test as they show.
 
With CPU in, roms out, fire it up to the garbage screen. Try toggling the slam switch on the coin door. If the screen changes, the reset line is probably working. If not, its probably not working.

Do you have the 14 pin plug connected on the daughterboard?
 
Yeah, I can toggle the screen a bit, does look like it "resets".
I always plug everything back in, even when testing, re-seated those connectors several times now.

Keep in mind this happened when the game was fully up and running, even played a game or two then just walked away from it and left it running and came back to this new issue.
Thinking something else died on the main board.

Gonna try and read through the Midway 8080 Processor Test Procedure guide a bit and see if anything clicks.
 
Thats the guide I used to fix my last one. If the slam switch is shifting around the screen, the cpu is running and the reset lines should be ok. (The slam switch IS a reset button basically).

I think at this point I'd yank all the roms out and get back to the bars again. Something definitely heated up and failed, could be capacitor failure. Theres 3-4 electrolytics near the power plug but I'm not sure if they can have any impact or not. I'd also go through the rams using the "ground pin 7" method to see if any make a major difference. If you have a probe you can also check pin 6 of each ram pair to see if there's pulsing. If one isn't pulsing you can start there.

(They may all NOT be pulsing if the cpu isn't doing anything, but if say most are but one pair isn't, then thats a problem.)
 
Last edited:
I think at this point I'd yank all the roms out and get back to the bars again. Something definitely heated up and failed, could be capacitor failure. Theres 3-4 electrolytics near the power plug but I'm not sure if they can have any impact or not. I'd also go through the rams using the "ground pin 7" method to see if any make a major difference. If you have a probe you can also check pin 6 of each ram pair to see if there's pulsing. If one isn't pulsing you can start there.

Grounded pin 7 of each RAM, no significant changes on screen. (more of a flicker if anything)
Pin 6 of each RAM appears to be pulsing, will double check as I'm still learning how to interpret this logic probe I have.

I read somewhere in that manual about the caps shorting out and possibly causing an issue, may double check/change those as well.
 
Logic probes are fairly easy to interpret - you have 1 (hi), 0 (lo), and pulse. 1 means its high, or basically 5v steady, low means the signal is low and is either very low volts or ground, and pulse means is pulsing with activity. No lights means no signal (or static I believe).

I've been cleaning up my work area for the past hour, ordered a few more bits and just came across this (of course AFTER I order) - going to see if they can add it.

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=A100662CT-ND

Its the friggin pins for that oddball power connector! Seems its called Tyco Duo-Tyne Flag Connectors!

Of course the pins are pretty expensive (technically) at a buck a pop, but for $7 you can restore your original connector, 3-4 bucks if you just do the power pins.
 
Logic probes are fairly easy to interpret - you have 1 (hi), 0 (lo), and pulse. 1 means its high, or basically 5v steady, low means the signal is low and is either very low volts or ground, and pulse means is pulsing with activity. No lights means no signal (or static I believe).

I've had one of these around for years, but never really used it. (Radio Shack Probe)
http://www.badassumptions.org/media/micronta-logic-probe-22-303.pdf
Now all of a sudden it seems to be stuck beeping and showing high when not even touching anything. (in TTL mode)

My biggest question is when trying to compare their o-scope readings in relationship to what a logic probe will show. Some things in the manual show a scope reading of say 1v, I would assume then that it should show LOW on the probe.
Like pins 17, 18 and 19 on the CPU should be LOW, I'm interpreting my probe as being high.
Also pins 8 and 10 on the E3 chip, 74LS74 should be LOW, I'm also seeing high there.

I'm getting a few things that are showing high and I would think should be low, but I'm also avoiding some of the sections that prescribe "cutting" out legs on chips, etc just for testing purposes. As some of these I do not have replacements on hand. (like the bus drivers - 8216's)
 
When the troubleshooting manual says "Check the Clock Driver IC 3245 output pins 2, 7, 10, and 15."
What am I looking for ? (high, low, pulse ?)
What about the inputs ?

Could I possibly get some actual voltage readings on the inputs and outputs ?
Inputs - 3, 6, 11, 14
Outputs - 2, 7, 10, 15

Pin 12 and 13 do read low as they should..... pin 12 goes high when slam/tilt switch is activated.
 
Last edited:
Figured I'd post what I have been looking at ALL day.

First one shows the pattern divided into four sections of garbage, two of which are in constant motion.
Second shows the blanking action when the slam/tilt switch is closed. When opened it goes back to the first.
 

Attachments

  • Problem 01.JPG
    Problem 01.JPG
    86.4 KB · Views: 19
  • Reset Active.jpg
    Reset Active.jpg
    53.6 KB · Views: 17
Back
Top Bottom