> > Several of the oscillator circuits are numerically unstable, and it was hard to get the > > solution process to work without making it so slow as to be useless. We also had to figure > > out how to emulate LM3900 op-amps more efficiently; they are a bit peculiar. > > You also had the issue of the datasheet misprint in one of the parts books and had > to search an earlier issue of a catalog to get exact values or connections iirc in > order to begin creating components before working on the Midway game. This is what > Derrick Renaud had predicted in his 2004 MW post when he gave his view of what issues > might be involved of emulating non-cpu games back at the time. Fortunate that this > work will enable other Midway 8080 racing games to have analog audio later.
Yes, the Motorola MC3340, where the new schematic had component values but a missing connection, while the old one had no values but correct connections (and also a couple extra diodes, one of which later turned out to make a difference in the chip's control response).
Incidentally, I think it's a good idea in general to check all the datasheet/manual versions of a part or system that you can, because you never know when you might come across discrepancies or changes like this.