Questions for my Squawk & Talk Bally folks! I have a Vector with a 100% repeatable issue that I am sure lives on the S&T. When two switches are hit in play, background sounds are silenced until they are brought back by a kicker being set off. All ROMs are freshly reprogrammed, all PIAs are working, as is the CPU (externally verified) and all chips reset back into their sockets properly. A cap kit was also performed by a previous owner (and looks OK). The switches work and score and this behavior preexisted on the stock -35MPU and continue with the AlltekSystems model. As this isn’t a switch error, I am trying to figure out which sound input line is failing here - all voices work, with a mysterious “Period Completed!” In the launch lane (strikes me as an error, too).

The 4049 IC tests good in my IC tester, all gates reporting in OK, and I am working to verify the 75LS155 demux. I'm wondering if one half of the demux is dead, and half still working, but I don't have any spares to prove this so I will order a few for fun.

Since I cannot know by inference what sound/voice is triggered by switch 35 in play (that left outlane) I can't know what line is shot in advance of just poking around and working every permutation. I have the same issue with switch 12, which is the rollover in the launch lane - the one that yells "Period Completed!" - what would it normally say?

I have a video of all this in demonstration, as well as a video of the switch matrix test (for the non-believers) and foolishly posted this to a FB group first, here:

https://www.facebook.com/groups/609432482409545/permalink/3069635316389237/

That link, if you can load it, will show you the videos I made, as I can't export them here yet (they're on my phone). The "advice" I got there was staggeringly useless, but well, it was an attempt to find someone with the experience (plenty of people with heart there, to be sure, including the ones trying to help).

Any ideas on how I can run this down?

demux (resized).jpg

AH, a hint! Yes, it reliably and always screams about an end of a period - that was my first clue. This helps!

I will check this when I get back to the board - I can verify that it was running stock 2532s, but perhaps the jumpers were messed up by the previous tech when the cap refresh was conducted...good idea to check again!

I cannot borrow anything from you until I repatriate you with some 2564s, which I now have, good friend. Say hello to your dear wife for me, and I will reach out to see if we can arrange some chip exchanges!

And, this is all I see that is "different:"

If speech is... ...then jumper this way
--------------- -----------------------------------
Required install J and remove H

Not required install H and remove J

Mine shows "not required" per the jumpers...I think this is incorrect for Vector??? Several others have been hit with solder, so this board was possibly in a different config before I got the machine.

Guys...so, under further inspection, I am using a 2518-61 board, not a 2518-61B model - and my schematic calls for the B model for Vector. I can verify that I do NOT have the three transistors off the speech module, Q3,4, and 5, in the upper right of the board. Could the earlier board model be my issue???

Yeah..research seems to show that a bare 2518-61 is used in a Flash Gordon. The A model was used in

Eight Ball Deluxe
Embryon
Fireball II
Fathom
Medusa
Eight Ball Champ
Beat The Clock

And the B model was used in

Centaur
Elektra
Vector
Rapid Fire
Mr. & Mrs. Pac-Man
Spectrum
Centaur II
Eight Ball Deluxe LE (Assumed)

And based on what I see in the schematics, they are indeed different in the speech sections, which is where my issue lies.

So, who needs a Flash Gordon Squawk and Talk?

This is EXACTY the mapping I was looking for - thank you! Yes, yes, all of my settings are exactly as you have described (Audit 18 at 03, EE jumpers removed, etc). With this mapping, though, I can now possibly spot a pattern to what I see in my game.

I have, burned them myself twice, and check the checksums twice, too.

Yes, I have all four - 2532s that program as they should and the jumpers reflect the config for them.

Yes, it does - all switches report and work as they should, hence, why I posted here. I wish it was simply a switch matrix issue...

I have - I have never seen a data issue like this - this isn't typical. I need to scan the input bus for what it sees, but was looking for a quicker route so I'm not on the bench scanning for bits and nybbles with my o-scope for hours.