Resurrected a dead Flight 2000 that came with two speech cards. Apparently the game had long suffered "no speech" issues, and given what I found the cards were not the problem.

One card is a Revision C dated 1980, and the other is Rev F from 1981. Visually they are nearly identical.

When I found the source issue, I connected the newer VSU card since it had a factory labeled F2K rom, and it worked! Played a few games and all seemed proper. Then I decided to try the older card (with an unlabeled ROM). And it worked even better! Because lacking a frame of reference, it turns out the first card was actually more... chipmunky? Fast, high pitched... and during the self test ("1-2-3-4-5-etc etc") it was trailing off with faint gibberish after the "5...", such that I didn't even know it was supposed to be saying words until I put the other card in, which boomed out a consistent, deeper monotone robot voice with added words.

My kids even joked the original card sounded like "splat" instead of "blast", ha!

Anyway, I'd already ordered the kits to recap both VSU cards (and sound board) per my normal rehab procedures, but they're not here yet. Meanwhile, I'm thinking a higher pitch that increases while trailing off, might in fact indicate an actual (as opposed to "assumed") cap problem: caps losing capacity and being drained before the full sentence is spoken...? So is that indeed the case, or does that particular card have another issue? It seems these cards are somewhat enigmatic with little troubleshooting info out there (they either work or they don't, and the unobtanium synthesis chip is often a culprit).

Any idea on the going value of these cards? Provided I get both working the same, I'm torn between keeping the extra as a spare or selling to recoup some expense.

I'll touch up the headers when the cap kits arrived. I'd looked at them previously and didn't note anything obviously amiss, but it won't hurt to check again when the iron is hot. If that doesn't equalize things I'll try swapping the synth chips... as pricey as those are don't want to push my luck.

BTW I have your updated code roms (v. as-of-a-couple-weeks-ago) ready for once I get this game dialed in. Thinking of calling it "Flight 2020" and modding the backglass with the year, and masks on the escapees Had I thought of it sooner, should've changed the inserts so instead of spelling "LAST BFFO" you get "COVD 2020" (or "BLAST OFF" becomes "COVID -19")... Anyway, I've been following your progress on the update thread and am eager to hear the finished speech updates too!

Yeah, the "2000" reference is such a ret-corny sci-fi throwback cliche... Sometimes I wish I had an old Gateway computer with the original "Gateway2000" logo. But I see what you did there! Don't forget, our space refugee / funereal vikings could be drifting amongst the Stars in a many a Galaxy; perhaps they'll be spotted by a Star Gazer and rescued by Quicksilver before decaying into a captured Freefall orbit...

...oof

barakandl Yeah I think I read somewhere that Meteor was an early licensed game to compete with Black Hole. I've never seen the movie but I bet it ticks all the 70's cheese boxes

Cap kits should be here in a few more days.

You might be on to something, as I noticed the 555 had been replaced on one of them (I don't recall which, but it might have been the first as I recall thinking "hey, if this has been fixed already, maybe I should use it first"... But I might have talked myself out of that. Ugh, can't remember and not at home now... But I'll check and keep this in mind!)

To me that's the most fascinating thing about the speech in these Stern VSU games. To my understanding, WMS games use compressed speech samples, whereas the synthesis chip in the Stern VSU generates a live analog synthesized speech pattern. The mildly defective VSU I have proves this to amazing effect. The one I have working better is truly crisp and distinct, and to me sounds even better than my Firepower (say nothing of later, abysmally-compressed samples like on Dr. Who, ha ha).

I applaud the effort to resurrect these games and replace these now-unobtanium cards with sample-based equivalents, but the live, analog synth aspect is truly amazing and worth preserving. It makes these games unique in a lineup. To think that any given VSU might sound "different" from another for any number of reasons, just like a human's own vocal processing, is awesome and just a little bit freaky.

Update: the Chipmunk card is the one with the replaced 555... (so indeed, you might be onto something)! Next question: "C27" is one of those tiny guys that looks like a glass diode. Do you recall what its spec is? I don't recall the schematic listing the components. Even more to the point, those types of caps are rarely supposed to fail, aren't they?

I'm kind of eager to see how this turns out though. Cap kits (less c27 anyway) should finally be here today. But I suppose if there's interest I could (should?) record before and after samples of the pitch comparisons on the cards.

I sadly no longer have the functioning bionic optics required to discern any markings on the cap itself. But it looks like an OEM installation to me...

20200611_131442 (resized).jpg20200611_131517 (resized).jpg

Hmm, I took another look at the card and you're right, there's lots of OEM solder joints that look unusually suspect. I'll give the card another pass...

But I'd already recapped and found some interesting results! Here's a video of three (yes, three ) distinct speech pitch patterns! Sorry about the length; my daughter and I wing these largely unscripted, and usual interruptions and unexpected chaos don't help (I seriously lack the time / equipment / household cooperation to do videos properly). But I timestamped the pitch jumps on the youtube page itself.

Oh carp... you're right, and I didn't misspeak: I used a 470 uF - or Microfarad - cap because I was so excited to finally see the numeric value and relative size in a little baggie I bought from GPE years back, I didn't register the discrepancy! Stupid brain.

Well, at least it's not the worst "educational error" one could make. In fact I'm kind of amused now: it'd be fun to socket that location and play with all kinds of values for that cap wouldn't it? Don't worry, I know better than to push my luck here.

Sadly I'm almost 100% certain I don't have any pF caps that value, but will of course double-check later.

This is my first/only Bally/Stern 2718-and-friends game, so no spares of anything there. But I have a junk WMS Sys6 driver and Sys9 MPU... I'll look there.

That's awesome of you! I'll follow up if I don't find anything.

So this almost qualifies as a necro-post, but I finally got around to playing with this again, so here's the follow up:

I wanted to try replacing the 555 as Andrew suggested earlier, but turned out I didn't have any. Finally got one and swapped it and... the chipmunking is gone! The pitch is still a little bit "higher" than my other card BUT the cadence is more in time. Put another way, maybe one card is a baritone and the other is a tenor, or something... I would say are passable at any rate. Pretty interesting to compare nonetheless. But I hate "the chipmunks" so I'm glad that's solved