Space Invaders Board Problems

ForKChop

New member
Joined
Jun 3, 2004
Messages
58
Reaction score
0
Location
Vancouver, Washington
Trying to resurrect a Midway Space Invaders.

I have tested the power supply as far as I can using the http://www.elektronforge.com/TestMdwPS.htm guide.
It passes both the beginning and intermediate tests, though the 18v line is actually closer to 15.3v. I do not have an oscilloscope, so I stopped testing there.

With the ROM chips installed it makes sound at startup and has horizontal strings of ascii garbage data, mostly, but not exclusively in the top row.

I attempted to test the ram as described at http://www.brentradio.com/images/SpaceInvaders/midway_8080_tech.txt
With all ROM chips removed, it shows no life and the screen is blank. I removed and reseated the 8080 without change. Again, no scope, so I believe the next step is to replace the 8080 chip and socket. Does this sound correct? Any other ideas of what I should try next?

Thank you for any help!
Vanessa
 
Progress.

Reset voltage at power supply is 23mV. Not sure if there's another spot to test it on the board.

Decided to measure and adjust power supply for voltages at the main board connector instead of at the supply. Reseated the 8080, and now with the roms pulled I have the vertical (really horizontal in this position) lines on the monitor. There is garbage data in there, on one set of lines it flickers away relatively quickly, but on the other it stays longer. Additionally, that's on static data, I'm sure as the video is written to the rams it gets all screwed up. So bad ram for sure. Possibly roms as well? I was thinking about ordering the Braze multikit, which should both a) bypass the roms and b) help pinpoint the faulty ram. Good idea? Any thoughts on how many ram chips I should order - I assume I should socket them as they are replaced?

Thanks again
Vanessa
 
Last edited:
Does sound like RAM issue to me off hand. Defintiely want to make sure the RAM is all good before troubleshooting ROMs.
Try grounding pin 7 of each RAM again and see if the lines clear up or change any at all. (no ROMs installed) Just keep in mind that that will only show you what "pair" of RAMs may be suspect.

You might try burning one of these test ROMs and see if it shows you anything.
A suspect RAM will show in the center of the screen as a letter or a number, then match that letter/number to one in the included diagram.
 

Attachments

  • inv_test.zip
    99.5 KB · Views: 29
An update. I received the RAM chips on Saturday and spent a good portion of the evening replacing all 16 RAM chips with sockets and new RAM.

Unfortunately all is not well. With ROMs removed, the lines are steady, until they switch, at which point there is random garbage for a few seconds until it disappears and everything looks solid. This is very similar behavior to before I replaced the RAM. First two pictures are examples of the garbage I see directly after it switches line sets.

With the ROMs in it seems to display random behavior, sometimes alphabetic garbage, sometimes just dots, sometimes words from space invaders or recognizable sprites. Occasionally it goes to a full screen of static garbage, as seen in the last picture.

A couple questions then, with the ROMs removed, and brand new ram, is that behavior normal? If not, any ideas of the cause? With new RAM, should I assume that the ROMs are also bad? If I get a Braze multikit is there a relatively good likelihood that it will work? I want to get this working, but I also don't want to keep throwing money at it with no idea of whether it will work in the future.

I don't have a programmer, so I don't have a way to burn the SI test rom. No oscilloscope either. Anything I can test with a multimeter I would be happy to do so.

Thanks for any help you can provide.
Vanessa
 

Attachments

  • garbage1.JPG
    garbage1.JPG
    65.7 KB · Views: 76
  • garbage2.JPG
    garbage2.JPG
    61.3 KB · Views: 60
  • romgarbage.JPG
    romgarbage.JPG
    95.8 KB · Views: 69
space invaders

Hello,
I have seem to had the same problem, the reset voltage is too high and should be 0.05mv ,mind was 0.27 until I changed the 2n4123 and 2n4125 in the reset circuit. It seems your board is not booting, ,how is your voltage...+5 -5 +12 all there, when measuring voltage all all 3 points use your digital meter only and switch to ac and each point should be 50mv or less.
 
space invaders

hello,
not yet ,I ordered a 8080a cpu ,but my power supply is right now ,try those 2 transistors and you have 104z (3) also may need to be looked at?do a cap change ,I did but it tuned out to be those transistors.
 
I don't understand why I should change the transistors. My reset voltage is in line with the elektron forge testing guide. Is that incorrect?

What does 104z(3) mean? Is this still in reference to the power board?
 
space invaders

Hello,
This is all in the power supply,you say the reset value is 23mv ,it should be 0.05 those 2 transistors are in the reset circuit ,look at those part numbers. the 104z is a replacement u can use for 0.1 50v there is 3 if needed .Just to verify you are measuring the reset value at in4732 (zener)cathode side
 
Hello,
This is all in the power supply,you say the reset value is 23mv ,it should be 0.05 those 2 transistors are in the reset circuit ,look at those part numbers. the 104z is a replacement u can use for 0.1 50v there is 3 if needed .Just to verify you are measuring the reset value at in4732 (zener)cathode side

0.05 what? Millivolts or Volts? I just rechecked, it's steady at 23mV / 0.023V. The elektronforge site says it needs to be 0.5V or less, is that incorrect? I am measuring it at the test point pictured here: http://www.elektronforge.com/Images/RESETTP.jpg

I've found the transistors, just want to be sure I'm clear on what it should be measuring.

Thank you
Vanessa
 
space invaders

sorry fat finger,my reset line is .005v and yours is .023v I would try to get it to .005, mine was at .027 before changing the transistors, just a thought ,let me give you the measurements from mine 2n4123 e =0 b=.728 c=.005.......2n4125 e=13.40 b=12.67 c=13.37 ,I would say if you have a 8080a to try? the caps .01 is another one to try ,tomorrow my 8080a should be here and will let you know if I have any luck
 
Last edited:
Shouldn't matter.

Looking at the 8080A data sheet the Input Low Voltage is minimum Vss-1 (-1v) and maximum Vss+.8 (.8 volts)

The input high voltage is minimum 3.3v and maximum Vcc+1 (6v)

So your logic low on reset can be as high as .8v and still be OK.

What is more important is that the RESET line must be active for at least 3 clock periods to reset the CPU. Make sure the power on reset circuit is doing that.
 
Shouldn't matter.

Looking at the 8080A data sheet the Input Low Voltage is minimum Vss-1 (-1v) and maximum Vss+.8 (.8 volts)

The input high voltage is minimum 3.3v and maximum Vcc+1 (6v)

So your logic low on reset can be as high as .8v and still be OK.

What is more important is that the RESET line must be active for at least 3 clock periods to reset the CPU. Make sure the power on reset circuit is doing that.

I'm assuming there is no way to check the reset line without an oscilloscope?

Thank you for the help on the voltage margins
 
Alright, doesn't look like it was the reset circuit. After channelmatic posted, I decided to read some of his repair logs and see if there was anything about space invaders. Troubleshooting 101. I re-examined the boards, and in one area where there was some scraping damage on the PCB one of the traces was broken. Not enough that it was easily visible, but once checked the trace for continuity and checked again with a loupe, there it was... a broken trace in the RAM part of the board. I repaired the trace, and lo the board boots!

No sound, but clean clear attract screen.

So a heartfelt thank you! Now to the other issues :D
 
WOOHOO!

Glad my logs helped ya :)

It feels great to bring a board back to life, doesn't it?

That 101 stuff is pretty important. I have an Area 51/Max Force combo on the bench and it was stuck in reset... Looked it all over and even the bent pins on the custom GPU were OK. It was a corner pin on one of the EPROMs that wasn't in the socket but hanging out over the short edge. DOH! Fixed the bent pin and it booted right up.

You'd be surprised at all the fixes that are simply gouged traces, bent pins, broken caps (especially in the reset circuit!!!) and other things that are obvious to the eye if you look closely.

As for the reset line, you can check it on power up with a logic probe. The logic levels will start out with active reset then quickly jump to the other logic state. Don't blink or you'll miss it.
 
Okay new question. That didn't take long did it?

I cannot add credits. The switch and all the wiring back to the pins on the board seem correct and have the correct continuity when the coin switch is engaged. It has the 5V on one leg, but no credits can be added. Is this a common problem, or should I start tracing back through the circuit?

No sound - looks like I need to replace the 1N4004 diode. I had already replaced the 1000uF cap to get the voltage.

Thanks again
Vanessa
 
Here's where a logic probe comes in handy. You put it on the board where the wiring harness connects - in particular on the input you wish to test, in this case one of the coin up lines. Trip the switch and see if the line momentarily changes logic states.

I'd have to look at the schematics for an SI but for standard JAMMA the pins are held high and when you push the switch they are grounded. If the pins aren't held high then there's a board issue. If they aren't grounding then it's a switch or wiring harness issue. If they are stuck grounded then it could be wiring, switch, or a short in the inputs on the logic board.

They aren't hard to troubleshoot if you have the wondrous little logic probe.

RJ

PS: Get one with audio capabilities so you don't have to keep staring at the LEDs. :)
 
Back
Top Bottom