There’s not much information online about these boards, and mine didn’t work when I got my Spring Break. Here are some notes. These are from memory and so will probably contain errors. Unfortunately I don’t own any of these games a the moment.
I don’t know if these boards don’t break much, or if the production of these games was so low that nobody cares, or both.
I worked on a few games that use an MA-766 board. There’s a later board, MA-886, that omits one of the test switches and one of the ROM slots, according to pinwiki.com. These board are reportedly similar (I don’t have any later 80B games to check with).
These boards have four ways to generate sound:
- H4 is an AY-3-8912, a synthesizer in General Instrument’s well-known sound chip family.
- Variants of these chips were used in a lot of ’80s home computers, the Intellivision, and the Atari ST, as well as Bally’s Squak and Talk board.
- K4 is also an AY-3-8912.
- E4 is an SP0250, a speech generator.
- The output of this chip is pre-amplified by the opamp at B1. The schematic indicates that R26 will be there “if used”; that is, it’s a jumper and presumably if the sound chip at E4 is omitted, R26 will also be left out.
- I don’t know if any games omit E4 or R26. Genesis, Monte Carlo, and Spring Break (at least) all have R26 and E4 installed. (Gold Wings also has this chip, labeled as “Orator”?)
- This chip is a relative of the SP-0256 which was used in the Intellivision Intellivoice (speech add-on). The Intellivision had precious little memory, so samples were pretty limited. That seems to have held true on Premier’s titles as well.
- E2 is a AD7528J DAC.
Each of the AY-3-8912s has three voices, so that’s a total of eight channels, six for music, one for noise, one for sound samples. That’s a lot!
Responsibility for controlling these is split up as follows:
- The E2 DAC is driven by he R6502-13 N1.
- This processor uses RAM H2 and ROM K2.
- All other sound chips (H4, K4, E4) are driven by the 6502 at T3.
- This processor uses RAM H3 and ROMs at N3 and K3.
Williams’ early solid-state games, including their video games, have a very distinct sound system because they wired a CPU to a DAC and used that to make noise. Some of the noises that come out of these 80B games are similar, likely because they, too, mated a microprocessor and a DAC.
Each of the four sound chips (H4, K4, E4, E2) are mixed by an opamp at B1. The inputs are mixed into 13, and the output is pin 14. The output of the sound board goes back to the auxiliary sound board, where it is amplified; then, to the volume control, where it is unamplified some amount, and finally to each of the two speakers.
System 80 games typically have four or five lines for picking which sound to make. This sound board actually has eight input lines, and a check of the schematic will reveal that no more than 5 are actually wired. These are pins 1 to 8 on A6P1, the only plug on the sound board. They are not in order. Eight sound lines could mean up to 256 sounds, but actually, all 0 means no sound. And since most games actually only use 5 sound lines, there are only 31 sounds. All inputs high means “off”. When any sound line goes low, this causes an IRQ on both CPUs. (Actually, there’s a little dance between the IRQ and NMI lines and a couple latches that I haven’t bothered to completely understand, but they are callouts to the programs in the separate ROMs.)
Most of the logic, including seemingly all of the sound generators, runs on +5V (comes from the skinny regulator board). The opamps, H1 and B1, are LM324s. These are connected to +12V DC and -12V DC, which comes from the aux power supply. Note that this is derived from the solenoid voltage, so if the solenoid voltage isn’t there, the amplifiers won’t work.
The (Hopefully) Blinking Red LED
The sound board has exactly one LED. In normal operation, this flashes on and off on about a one second interval. This light is software controlled. The processor at T3 can select it by setting A13=L A14=H A15=L (T4 ABC=LHL, respectively, on pins 1, 2, 3) to address the G3 flip-flop, and setting data bus D1 low to turn on, and high to turn off. The current value of the latch is stored in the flip-flop G3, an 74LS374. G3 is part of the apparent chip select logic. It can pick either one of the AY-3-8912s or the SP0250. So presumably the CPU should hit G3 pretty regularly. Note that the processor at N1 has no control over the LED.
SW1 is an input to the 6502 at T3. It does not talk to the 6502 at N1, which won’t be tested. (Pressing this should emit a beep.)
SW2 does a reset of both 6502s and both AY-8912-3s. It’s a good honest reset.
On my Monte Carlo, there’s a “beep” when the game powers on. After a few seconds, the LED starts blinking at about a slow, 1-second on, 1-second off pattern. Pressing the test button emits a pleasant tone.
On Genesis, there’s a “beep” when the game powers on. (I haven’t verified further.)
Try swapping chips from a known good board. Sorry, that’s what worked for me.
This board can be analyzed by “divide and conquer”, particularly since there are two sides running more or less independently. I would expect that if it makes no noise, the amplifier is shot, or something is screwed up in the sound select circuitry. If only some of the sounds are missing, it should be possible to figure out which half of the board didn’t do its job and work backwards.
When Things Don’t Work
No sound unless volume is at max, and it sounds terrible.
Check the solenoid voltage. I had pulled the solenoid fuse so that didn’t have to listen to the drop targets reset every time I turned the power off and on. No solenoid voltage meant no voltage for amplifiers.
Blinking red light is dim. Sounds are mostly static.
When I got my Spring Break, it didn’t make noise. Actually it made static, and plenty of it. The problem turned out to be a bad ROM (which, I forget). I proved this by swapping chips out of a known good board. So, if you buy one System 80 game, buy a second one so that you can test boards in it.
The red light was also not blinking, an indication that the N1 CPU side of the board is at fault.
Chris ran into a problem where several sound boards, including some known-good, did not work in a Raven. The problem was the external reset line was being held low by the display unit. As a result, the CPUs on the board couldn’t boot.
The solution was to disconnect and troubleshoot the display.
No Sound/Wrong Sound/Slow Sound
Chris had a Gold Wings in the shop and the sound “didn’t work”. We powered it on and found that:
- if you pressed the test button, it would beep, indicating good input voltages
- we did hear the game play music once, but not consistent with the rest of the game, and it usually didn’t work
- the sound would slow down as the game warmed up
This game is in rough shape and had corrosion on the connectors. That helped somewhat.
Swapping in an alternate driver board showed that something was wrong with the sound-select scheme, so the last buffer chip on the driver board was replaced.
Finally, the scope showed that the clock worked, but the divider chip that takes the clock output didn’t work. Looks like a heat problem. Replacing the 74LS04 fixed the sound issue.
Also note that a bad volume control, bad speaker, or bad sound amplifier may be the actual problems.
This board gets a reset signal from the MPU and the display. That means a MPU or display holding the external reset line low will prevent the sound board from booting.
Getting ROMs for these games is difficult. Pinball Resource is the official supplier, and the chips aren’t that cheap.
The AY-3-8912 chips are uncommon. Wikipedia suggests that there are some that are legitimate, but incompatible, because they have different baked-in ID numbers. Hope you don’t need one.