Space Invader woes

Sid723

Member
Joined
Jan 16, 2011
Messages
92
Reaction score
2
Location
Mokena, Illinois
My SI stopped working after years of service. No pic, no sound, nada. 5, 12 and -5 are all good on board. In t/s I found no clock signal on CPU. I replaced the 7404 and the 2 330ohm resistors are good and the 180pf is good. I also replaced the 20MHZ crystal.

Anyone have an SI that can help?
 
What's "t/s"?

There are a number of ICs between the oscillator and the CPU clock input.
Grab a copy of the schematics and follow along: http://www.brentradio.com/images/Other/Docs/SpaceInvaders/MidwaySI_Schematics.zip

From the 74S04 @ A7, it goes to a "9310" @ C7. That's eqiv to a 74160 (decade counter). So use your logic probe or scope to see if its outputs on pins 11-14 all look good.

If so, continue to the LS42 decoder @ C6. If it's inputs are cycling properly, each of its 10 outputs should also be pulsing.

From there, it goes thru a couple of flip-flops (B5 & B6); check their outputs for pulses.

Finally, there's the 3245 clock driver chip @ C5.
 
"t/s" means troubleshoot. The 74S04 I replaced with a 74LS04 (all I had handy). I'm not sure if that would make a difference or not. The 74160 was also replaced. The 20MHz crystal was replaced, but I am not sure if it was an exact same type. Anyway, when I check with a logic probe, I get no signal on the 74LS04 pins 3, 4, 5 or 6. And I do have all the original manual and schematics.

Can anyone verify if I should be seeing a strong signal on my logic probe?
 
soldier joints and traces

these games are old try reflowing the soldier , or look for cracked solder joints,also the mold midway series games that have that pcb the edge conectors get bad or this apply soldier to the traces on the edge conector, may help, or may not, come across this often, also did you check your voltages with the pcb pluged in cause it could be a weak 2n3055 seen that a lot, when load is aplied voltage drops, just some tips.
 
The 74S04 I replaced with a 74LS04 (all I had handy). I'm not sure if that would make a difference or not.

Actually, that is a pretty big difference. A TTL "S" series should only be replaced with an exact "S" or a "F" series.

Edward
 
OK. I replaced the 74LS04 with another and I got clock signal at pins 15 and 22 on the 8080. Sometimes I get noise and some numbers on the screen. I will check proms next. May even have a bad socket somewhere too.

I'll update later.
 
Last edited:
The conventional trouble shooting technique for Midway 8080 boards (as first written in the "Standardized Test Procedure") is to remove all EPROMs and check for "good vertical lines". Note that they're only vertical when the monitor is viewed in the conventional orientation... and it's NOT installed in conventional orientation in SI... (so they will appear horizontal)

The two DOA SI boardsets I worked on, I started by doing this, and once any issues were resolved (and it was displaying "good vertical lines"), I then proceeded with the SI test ROM: http://www.outerworldarcade.com/arcade/space_invaders/space_invaders_test_rom.html
 

Attachments

  • good_lines.jpg
    good_lines.jpg
    33 KB · Views: 53
That is cool. I didn't know that SI had some support like that. Anyway it can be modified to place it in one of the open slots? Maybe even have it run when a dip switch combo is seen on reset? That would be handy.
 
Update: All ROMs are good. All power is good. I now have the boards on my test bench and I found the reset (pin 12 on cpu) is high (5volts). Schem has it going to pin 16 of daughter board connector and pin 10 of power connector (although nothing is at this pin).

On daughter board E4 & E5 pin 1 (reset) are both low. This comes from F3 (7404 inverter) pin 10. Not sure where pin 11 is going, but connector for speaker power says "POWER ON RESET". I try to bring pin 6 to ground, but reset shows a glitch, then stays in reset.

Anyone have a clue?
 
Ok reset is good. When I put boards back into game, the reset on the cpu goes low and pin 6 of speaker power (reset in) goes low too. When I had bard on test bench, I used pin 1 to bring pin 6 to ground. Not really true ground on board, pin 2 was better.

Anyway, now I can't read fast activity on the address or data lines. They do change, but seems like they are very, very slow. For instance: AD1 on cpu will stay low for about a second then go high. I don't think my logic probe is bad, because I used it previously to t/s a MsPac board. And when I check pin 15 & 22 of the cpu, I get a fast changing signal.

I left it on for a while. Every couple of seconds it makes different sounds.

Any more ideas?
Anyone ever see this before?
 
Ok, I found out my problem once again was a bad A7 (74S04).
Now I finally get good signals on my Address and data lines.

Only problem is game makes no sound or picture.

I will need to do further t/s on my test bench.

Does anyone have any idea where I should start?

It would also be nice to have a small monitor to put on my bench to let me know if I am getting a picture.

Any ideas would be appreciated.
 

Attachments

  • si_clock_ckt.jpg
    si_clock_ckt.jpg
    67 KB · Views: 29
Last edited:
Test bench observations:
RAMDA7, 4, and 1 on C4 (74166) all stay low. This could be be a sign of bad RAM chips.
I started down this path by looking for what generates composit picture and traced it to the pin 13 of C4 and pin 8 of A6. I get signals from pin 8 of A6, but not from C4. Over by A6 there are a lot of timing signals, so I assume that A6 is generating the composite picture sync signals, while C4 is responsible for picture information.

Which at this point I am lacking. Next I will see if I have some RAM chips that I can replace and see if the output for that ram starts working again.

Till next time.
 
I recently repairs a friend's SI arcade machine.
I ran across 4 bad caps on the sound pcb board. This is not the first time I have run across bad caps.
You might want to use an esr meter check if the caps are good..
 
Hey guys. I did some more troubleshooting (t/s) and I found something very interesting indeed:

Since I was getting no picture and I checked the output of C4 (74166) and found that some of the signals from the RAM was not getting there (DA7, DA4 and DA1).

Tracing this back to the RAMS, I found that none of the RAMS were giving me an output at all (pin 7 of each ram).

I was getting the address and data signals going to the RAMs, so I checked the signal for write enable (RAM WE) and found that all of them were always low!! That makes sense now.
Since all the RAMs are being written to, then the outputs will be disabled (in theory anyway). So, this is why I get nothing on the outputs of the RAMs.

I then traced the WE signal back to pin 9 of F4 (74157 Quad 2-line to 1-line data selector/multiplexer). According to the truth table, pin 9 should go high when:
1. pin 11 is high and pin 1 is low
and
2. pin 10 is high and pin 1 is high.

I checked for these conditions on my dual o-scope and they do happen, so the pin 9 should have gone high at some time. Checking pin 9 on the o-scope and found it never goes high.

I think I have a bad 74157 chip (F4).

So, now the search begins for a 74157. That's a whole lot better than looking for these
antiquated RAM chips.

Wish me luck. Till next time.
 
Last edited:
Eureka!

That was the problem. The 74157 (F4) chip was bad.

I just happened to have a 74LS157 in my parts bin. I wasn't sure if it would work, so I put a socket in the board first, then put the chip in.

That did the trick. Looks like the bad output (always low) was causing the RAM chips to go into write mode and kept the outputs off.

Well wrap up another case of the Space Invader woes.

Thanks for your input guys.

Have a great one!!
 
Good work on your part, thinking through what you saw and isolating the problem IC.
 
Thanks for the great troubleshooting log! Nicely done!

Congrats on bringing a classic back to life!
 
"Thanks for the great troubleshooting log! Nicely done!

Congrats on bringing a classic back to life!"

I think it would have been better if I showed short diagrams of the circuits I was investigating at each step.

Maybe I will edit this later and add pictures in.

I am happy to report that my old Space Invader is still alive and kicking.
Just played it again today. The gradson is grateful too.

Later Dudes!!
 
Last edited:
I get a black screen on my SI
5v and 12v are all good
I am new to repairing boards so I think I should start with checking the clock but unsure exactly what I should test with my logic probe. According the schematic it looks like pin 4 on the A7 chip but I could be completely wrong

Can someone tell me how to read that clock area on the schematic and what should test what with the logic probe

Thanks
 
I get a black screen on my SI
5v and 12v are all good

How about your -5V supply? I seem to remember that the RAM (and maybe a few other ICs) on SI is tri-voltage (required +5V, +12V and -5V).
 
Back
Top Bottom