Space Invaders Deluxe

KevinMullins

Active member
Joined
May 27, 2009
Messages
2,187
Reaction score
10
Location
Mcloud, Oklahoma
Been doing a ton of reading on this, just need a little clarification....

It originally had TMS2716 ROMs.
I've since burned a new set of regular 2716's.
Power supply checks out good.
I do have two suspected bad RAMs, waiting on parts order.
I don't have another 8080 CPU handy to test that out yet.
Test ROM just brings up garbage. Slam switch does effect it, just nothing note worthy.

A. Can someone verify I restrapped this one for 2716's correctly ?

B. From my reading it sounds as if the lines I have on screen are telling me it's still not reading the ROMs. (or some portion of them)
 

Attachments

  • 2716 Strapping.jpg
    2716 Strapping.jpg
    96.2 KB · Views: 62
  • SID.JPG
    SID.JPG
    55.9 KB · Views: 62
Voltages are correct to my 2716 ROMs.
Pin 12 - GRD
Pin 24 - 5v
Pin 21 - 5v

Need to double check the address line strapping I guess.

(on a side note if you happen to short pins 21 &20, it'll let you know....the audio goes nuts, scared the crap out of me with my head all up in the cabinet) :)
 
Ok, so after studying the schematics a bit more and using the ROM datasheet and probing each solder point for power and continuity to the ROMs....it actually all started to make some sense. Maybe it's this revision of board or differences between regular SI and SID, but I don't know WHY most of the internet references has stuff crisscrossing all over the place for this particular restrapping job.

S1 = ROM Pin 22
S2 = ROM Pin 21
S3 = ROM Pin 19
S5 = ROM Pin 20

Each potential solder point is either an address line, a voltage, or ground. You're redirecting what you want to go to that particular pin on the ROM. Verified ROM pinout using datasheet of course. Easy enough....
 

Attachments

  • 2716 Strapping A2.JPG
    2716 Strapping A2.JPG
    28.8 KB · Views: 39
  • 2716a.JPG
    2716a.JPG
    46.2 KB · Views: 36
  • SAM_1011.jpg
    SAM_1011.jpg
    94.5 KB · Views: 43
Oh.... no better results though.
Same image most of the time, consistent lines.
Every once in awhile I can turn it off and back on and get some random garbage or letters to show.
 

Attachments

  • SAM_1012.JPG
    SAM_1012.JPG
    70.6 KB · Views: 43
Last edited:
Just gonna talk to myself for a moment.... :)
(notes)

ROM Pin 18 (CE- Chip Enable) is being held high, which would have it put in Standby/Power Down mode if I'm reading the datasheets correctly.

Pin 18 currently 3-3.3v
Vil = -0.1v - 0.8v
Vih = 2.2 - Vcc+1 (6v)

So the ROMs aren't even trying to run.
ROM Pin 20 (OE - Output Enable) is low though like it should be.

Hmmmm.....
 
Last edited:
More notes...

E2 - DM74LS42N (BCD/Decimal Decoder)

BCD Inputs Decimal Outputs
D C B A 0 1 2 3 4 5 6 7 8 9
L L H L H H L H H H H H H H

My understanding is all the inputs should be high on E2 in order to make all the outputs LOW which is what the ROMs want on pin 18. (?)

This leads me into the address bus area where I tend to get lost....but I guess I'll poke around it some more.
CPU and/or RAM ?
 
For what it's worth - I saw the same thing on a board I was working on, using a 2716 test rom. I tried grounding pin 18, as well as connecting it to pin (I think) 20 which was low at the time, and then even put an extra socket in between and left pin 18 floating. It didn't make a difference. I actually got the board working (whoda thunk), and went back and tried pin 18 in all different ways, and it seems like nothing made a difference.

Long story short, it seems despite what you read (which is prob what I read), that it doesn't matter that pin 18 is high.

Once I got it to boot from the rom, I was getting garbagey screens with a black box with most of a letter inside. I believe this was the test rom pointing out a bad ram.

I took a ram chip and tried piggybacking each chip one at a time, until I found one that seemed to change the partial letter to an "F". Left it on there, rebooted, and sure enough it actually came up.

(I'm not 100% sure this is really the case, but as I piggybacked the chips, another number (3) showed up at some point. Ironically I was on the third chip in the row farther away from the edge of the board. When I got to the 6th pair, the letter turned into F, and I was on the row closer to the edge. I think the test rom points out bad rams using letters for the row nearer to the edge, then numbers on the other row. Again whether or not this is the case, who knows, but it seems to make sense, and piggybacking chip "F" per the test rom got it to run!)

The test rom takes a few seconds to actually boot - it does a ram test to start which kinda flushes the screen a few times. When you turn it on, give it at least 5-10 seconds.
 
Last edited:
PS - Just went back and looked your jumpers - COMPLETELY different from what I have!!!

attachment.php


This is the way my board came, with originally having 9316B roms, supposedly compatible with regular 2716's. I desoldered S2 to try un-grounding and putting 5v on it (I think it goes to pin 21 on rom H), but it turns out that didn't matter either! Once my ram sockets come in I'll solder it back together.
 

Attachments

  • photo.jpg
    photo.jpg
    71.4 KB · Views: 281
Last but not least, I would yank all the roms and just use the test rom until you get that working. That eliminates all the other rom sockets.
 
Long story short, it seems despite what you read (which is prob what I read), that it doesn't matter that pin 18 is high.

Interesting.... only because I was going off the ROM datasheet which led me to believe that pin 18 had to be low for ROM read operations.
Guess I'll ignore it for now.
I have RAM on order and was going to change the two suspects I have first before digging to much further. Occasionally I would get the lined screen and it would have random dots on it, so I went ahead and checked the RAM by pulling the ROMs and grounding pin 7 on the RAM and such as described many times, found one pair that would clear it up.

Once I got it to boot from the rom, I was getting garbagey screens with a black box with most of a letter inside. I believe this was the test rom pointing out a bad ram. I took a ram chip and tried piggybacking each chip one at a time, until I found one that seemed to make a big difference on the screen. Left it on there, rebooted, and sure enough it actually came up.

Ok, glad you mentioned that.... because I have seen the same garbage or blocky lined screen with a random letter in it with the test ROM, but it didn't quite fit the description of what I was expecting. Most of the time I had a letter G, sometimes a letter E dead in the center of the screen. Now that you've described it, then the E code would match one of my suspect RAMs, but most of the time it showed a G, which does not. So I'll probably replace my previously suspected ones (E which is H12 and G12 which is it's "pair") and then try it again and see what it shows.
(attached pics show this)

The test rom takes a few seconds to actually boot - it does a ram test to start which kinda flushes the screen a few times.

So did you ever get the test ROM to come up to the test screen with OK RAMS, ports, sound test and all that ?
 

Attachments

  • SAM_1013.jpg
    SAM_1013.jpg
    97.1 KB · Views: 32
  • SAM_1014.jpg
    SAM_1014.jpg
    95.1 KB · Views: 25
Last edited:
PS - Just went back and looked your jumpers - COMPLETELY different from what I have!!!
This is the way my board came, with originally having 9316B roms, supposedly compatible with regular 2716's. I desoldered S2 to try un-grounding and putting 5v on it (I think it goes to pin 21 on rom H), but it turns out that didn't matter either! Once my ram sockets come in I'll solder it back together.

Crap... thought I had that all sorted, didn't want to think about it again.
Well it's not COMPLETELY different....S1 and S5 are the same. :)

But your jumper setup would give ROM pin 19 (S3) 5v and pin 21 (S2) grounded. (?)
And you have standard 2716's in yours now ? (not 9316, not TMS2716)

I read a bunch of threads here and there about 9316B compatibility, will have to go back through and re-read those. Some say it's "iffy", so I'm just wanting to clarify what you have.

EDIT:
Did some comparing and continuity checking and we actually have three points the same.... S1, S3 and S5.
Yours has S3 going up to S2's "A10" spot so to speak. Why they didn't just go down to the left like I did I dunno, checks out with the meter. I think the older boards didn't have the lower spot, so maybe it was factory habit to put it where yours is. (I actually had my S3 like yours originally)
So the only one we have different is S2, which yours is going to ground. I had set mine to go to the 5v as that's what the ROM datasheet says is required at pin 21 (Vpp) to READ. I'll take mine off and see what difference it makes, since on yours you mentioned it didn't seem to matter if it was connected or not to ground.
 

Attachments

  • Yours.jpg
    Yours.jpg
    72.7 KB · Views: 22
  • Mine.jpg
    Mine.jpg
    94 KB · Views: 21
Last edited:
Yup a standard 2716, originally 9316B, and I did nothing to change anything in the end.

Incidentally, the one I fixed is exact same part # board as well.

The problem I saw with that black box is that the bad rams will actually screw up the letter, but yeah its possible thats a g or an e. If your looking at the board with the daughterboard to the left and the ram chips across the back, G would be the 7th set from the right, rear row.

And if that black box is truely working properly, then chances are your rom is functional too since its doing the ram test.

To answer the last question - yes I got it to come up with RAM OK and it was cycling through the sound tests.
 
Been doing a ton of reading on this, just need a little clarification....

It originally had TMS2716 ROMs.
I've since burned a new set of regular 2716's.
Power supply checks out good.
I do have two suspected bad RAMs, waiting on parts order.
I don't have another 8080 CPU handy to test that out yet.
Test ROM just brings up garbage. Slam switch does effect it, just nothing note worthy.

A. Can someone verify I restrapped this one for 2716's correctly ?

B. From my reading it sounds as if the lines I have on screen are telling me it's still not reading the ROMs. (or some portion of them)

Your picture shows the proper strapping for a "regular" 2716 on that board revision.

There's only one jumper that is different from a "regular" 2716 and a 9316B...jumper S2/pin21. On a 9316B it is tied to ground. On a regular 2716 it is tied to +5VDC.

Now, I've personally had plenty of regular 2716 run just fine with pin21 tied to ground.....I've also had plenty NOT run with pin21 tied to ground, but run fine when pin21 is tied to +5VDC. I've never bothered to figure out which manufacturers 2716 need pin21 tied to +5VDC....and which one don't care. If I'm working on a board with 9316's( or whatever oddball ROM) and I'm swapping in known good 2716's....I'm spending the 10 seconds to change one jumper. This way I know it's one less riddle I have to figure out :)

Edward
 
orion3311 - Do you know off hand what brand/type 2716's you used ?
I've been using M5L2716K's and D2716's.

I still have about 4v to pin 21 with the S2 jumper taken out.
I also seem to have a better response from the test rom with S2 either disconnected or tied to 5v. Seemed to kinda lock up when tied to ground.

I'm going to hold what I have now and wait on the new ram to arrive and then square some of that away and then go from there. (gotta get the obvious out of the way)
I also ordered another 8080 CPU since I simply didn't have one handy to test with.
 
I think they're AMD, but I'm not 100% sure. They're pretty old (1976) but work fine.

My thought is...if that black box is coming up - its working. It means the program is running and doing the ram test. If it sees a bad ram...end of story thats all it'll do is display the black box.
 
Just waiting on the ram and CPU I ordered to show up.
I will most definitely follow up with it.

Now, I've personally had plenty of regular 2716 run just fine with pin21 tied to ground.....I've also had plenty NOT run with pin21 tied to ground, but run fine when pin21 is tied to +5VDC.

I did play with that just for a moment and mine seems to have better/more response when S2 is tied to 5v or left off versus tied to ground. With S2 left off I still had around 4v at pin 21 of the roms.
I'll know for sure once I get the test rom or game itself to boot properly.
 
Last edited:
(Sorry for the really long post but figured this might help someone troubleshooting these boards down the road. Again - I'm not an expert; just learning as I go, but man I'm starting to get it!)

I just modded an older standard SI board I had for the same strapping, and it also worked with the 2716 and S2 to ground.

I also verified that the black box is a ram test, and indeed it works as I detailed earlier. With the board sitting on the bench face up and the daughter board to the left (And the ram towards the back), the farther back row is H to A (left to right), and the other row closer to you is 8 to 1 (left to right). It does round out numbers like 8 so if it looks like an 8, it could be an eight or a 6, but not E. Bad rams can take out columns so the number or letter maybe missing a piece. For a quick test try piggybacking the ram on top of the suspect ram (making sure the pins make a tight fit).

Each column of ram from left to right runs off each data bus from D7 to D0 (left to right) off of the 74LS174 at D7. Each data bus is connected to each column through the 330 ohm resistors above each column to pin 6 of each pair/column (A column would be 8 and H, 7 and G, etc).

I just fixed this board (woohoo!!) - once I replaced socket H and the cpu socket, put in a fresh cpu, it came up with the inverted lines like it should. Modded the strapping (bridges were actual traces...made tiny cuts that could be reversed if need be) and the rom started working and came up with the black box and most of the number 8. Wasn't sure so I replaced E first, which obviously didn't work. Replaced 8 and still got same thing, although the missing row from the number 8 would fill in when grounding pin 7 of 8 or H, which practically verifies something in that pair is bad. Replaced H, still no dice.

Did some comparisons with the logic probe and found pin 6 (Data bus D7) was low when all others were pulsing. Traced that back to pin 14 at D7. Replaced D7 - still no dice. Finally looked at the traces again, and found out they continue from pin 14. Followed it and found a solder bridge on a capacitor. Fixed that, and board came up to the test rom! Woohoo again!

Board was randomly resetting because I had it sitting on its side with no brackets on the daughterboard. Turned it upright and re-seated the daughterboard and 14 pin plug, and seems rock solid now.

Also while going through a few daughterboards I found the 14 pin plug can get pretty dirty and make a bad connection, causing resets. That 14 pin plug is responsible for the reset line as well as the 18v for sound.

Last but not least, the diodes responsible for the 18v on the power supply (thank you Elutz it finally sank in) which Elutz said were compatible with a 1n4007 - well I found Radio Shack's multi-pack of diodes had 4-5 of them in it, so I was able to fix both power supply boards I had and got sound again.
 
Very cool.... thanks for the info and tips. I like the learn as you go approach myself. (otherwise I wouldn't know anything, know what I mean?)
Hoping my stuff comes in in the next day or so, ready to get back on this one.
 
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.
 
Back
Top Bottom