I haven noticed the -51 sound board running future spa software has at least one sound effect that will speed up or slow down depending on what 6802 chip is installed. The sound effect is the game start noise. Some chips it is very slow and takes 2 seconds to complete the effect. Other chips it takes 100ms or so to complete.

I have tried numerous different chips and STM, SG, Moto, Hitachi all will show some variances and I have not yet identified a pattern, but chips in the same lot codes will act all the same way. Every SG Thompson made in France are always super fast type for example. Hitachi I don't know how to decode their date scheme but usually these are "slow" type. Motorola varies with newer the date code more likely to be "fast type" but not always. Not enough data to tell a pattern yet.

As far as I can tell this only effects Future Spa's game start noise and maybe a few other sounds in an insignificant way. Has anyone else noticed any other -51 sound effects vary by what CPU chip installed? Any idea what is going on to make some chips go fast and others slow? Seems like there are some undocumented differences cpu chip to cpu chip depending on date code, manufacture plant, and/or brand.

Here is an example if you want to listen. I did not capture any "slow" ones.

"medium speed"

"super fast speed"

That is interesting. The PSG part of the S+T is pretty similar to -51.

Watching youtube videos of future spa older than the replacement sound board and I hear some variances in this game start noise. Some have the slow speed and others medium. Have to assume majority of original -51 sound boards are on the 6808/6810 setup, but internal ram vs external could be explored. Since some 6802 can be the slow type, I am not sure if that is going to be it, but I have a few tubes of 6808 and probably have an original -51 around somewhere to test with.

The issue with sound speed does not appear related to the internal RAM, at least in the -51 board. I can take a super fast 6802 with the internal memory disabled (using 6810) and it still is super fast.

All of the 6808s i tried where either medium or slow types, but they are all used pulls from Bally / WMS games when I swapped them out for 6802 so I could salvage the 6810 for another purpose. That means the 6808s are all date codes from late 70s and early 80s. While the 6802s I have that seem fast are mid 80s and 90s, except for Hitachi I don't know how their date code scheme is decoded. That said it could still be a 6808 vs 6802 thing, just not with internal RAM

Future Spa and maybe Hot Dogg'n where the only games that seemed to have issues with a sound speeding up. In hot dog its only the power up alert "tune" which is a mash of noises and doesn't really matter in my mind. All the other games I cannot find any sped up sound effects.

I guess Future Spa software is timing based on how long it takes the processor to complete certain things. In old date codes / certain batches it completes it in less cpu cycles.... guess.

Not the end of the world to have the game start noise be super fast in future spa, but annoyingly makes me sort through CPU chips to find "slow" ones. I mark the underside of the fast ones so I know and then use those elsewhere.

So far it only effects the bally -51 and maybe the squawk and talker. I have used 6802 in WMS square shaped sound boards and it has always been OK.

Quench pointed out by changing RAM values in emulation he can get the power on and start tune speed to change in Future Spa.

It seems like Bally does not format the RAM at boot up. Some RAM seems to be more likely to power up with data that makes the game start tune slow or fast. On the replacement -51 sound board I am using 6802s with internal RAM and chip to chip I get different results as far as how fast the power up tune is. All CPU chips with the same date codes typically behave the same though, all fast, all medium, or all slow. Currently sorting out and only using the 6802s that always default to "slow" or "medium" which is probably what Bally intended. Future Spa, Hot Dogg'r, and Frontier are effected. Have not found any problems in other games.

External 6810 RAM was typically always used by Bally from the factory. Not sure what to expect with external 6810 RAM, but I think you you would still get random future spa game start speed results with unless that RAM is more reliable to have the same data between power ups. Watching youtube videos older than my replacement sound board I can hear at least two different speeds.

It not a clock speed issue. That is where I went to check first.

I actually abandoned a 2mhz crystal for problems like that on the MPU. The 6802 is very happy to work with a 3.58mhz crystal, MC and EF6802 does not like 2.00mhz unless you really dial in the load capacitors and add a series resistor. HD6802 worked better at 2.00 for whatever reason. Complete external oscillator fixed all that.

Yes, this is the same kind of thing that bugged Meteor!

In meteor every time I installed PCD5101P the sound bug goes away. Other RAM brands, or mixing two brands, I would get the sound bug. I figured it was a speed thing since PCD5101 was the fastest. I think it was just luck that PCD5101 defaults to an expected '00' or whatever in the memory location that causes the sound problem when power is lost. The other ram that was assumed "too slow" was just more likely to corrupt in away the sound problem happens. There is zero way to clear the RAM on the board without special software. Then it was reported one brand of NVRAM had a sound bug, but mine did not. I turns out I was just lucky the RAM tester I use pattern left on the chip worked with Meteor. Alternatively I think you can fill with all zeros and meteor, and other MPU 200 games, sound right.

I have a Future Spa RAM clear test to try (THANK YOU). I will report back later today.

I would guess during Bally's testing of future spa they used 6810s which consistently default to the "slow" speed between power ups and they never realized there was a problem.

The modified ROM that clears the sound RAM at power on is producing consistent results, always "slow" future spa start tune no matter what CPU chip is used. "Slow" game start tune is what i think is what Bally intended, for sure not the super fast speed one. Mystery solved

thank you

you know what the sad joke moto did to you guys is... Every single processor with Bally's custom part number for a 6808 I have checked is actually a 6802 with functional internal memory. Same thing with Williams, yet both stuffed and used 6810 with them. Moto sales rep got a commission or something for unloading 6810s =D. Some of them you can see that they rubbed off the 6802 mark and put your custom bally part number down. Occasionally the 6802 mark can still be partially seen.

My reel of FM16W08 got pushed back again, f^&king cypress. I picked up some used FM1608 to get by from a seller that had been sending me nice, clean leg, socket pull, original FM1608B 2010-2011 dates.. The last batch they remarked FM1608B to the inferior part number FM1608. With the right light spectrum at the right angle you can see the original marks still. Ugh.. why downgrade the chip! I guess everyone searches for FM1608 and not the B version