Zaccaria Magic Castle Sound BoardThe repair information here is only intended for anybody who has electronics repair experience, and is only a guide. Aussie Pinball and I can take no responsibility if you cause serious damage to yourself or your machine by repairing the circuit board yourself. If you do not have experience working on circuit boards, then you do so at your own risk, and cannot hold myself or Aussie Pinball responsible for any potential problems or damage caused to yourself or your pinball machine.
I received a zaccaria Magic Castle sound board (1B11136/0) for repair recently and thought I would document the repair. There is alot of information on David Gersic's site
(
http://www.zaccaria-pinball.com) and have collated some of the information plus added my own findings.
I didn't have access to the pinball so the board had to be repaired and tested separately. A PC power supply and 4 ohm speaker can be used for bench testing. The sound board requires +5V, -5V, +12V to operate. If you are using this method you will need to connect the 5V rail directly to the cathode of the D2 IN5402 diode in order to supply the correct voltage to the board. In the actual pinball, the Zaccaria schematics show that power supply actually supplies 5.6V to the sound board. Once it reaches there, it passes through a protection diode mounted on the board which drops the voltage back to the normal 5 volts required. You also need to make sure that there is a 47 ohm resistor in series with the speaker or the TDA1510 amplifier chip will instantly be damaged.
First run a self test on the board using the red self test button. This will perform testing of the major components and play some speech but it does NOT test the music or sounds played by the AY-8910 and 1408 DAC ICs.
If the board completes the self test the LED will flash 5 times and then play some speech. If the test fails, then please use the following chart to determine the fault.
Self Test- Number of Flashes
1 - 6802 CPU RAM tested OK
2 - 6821 PIA 1
3 - 6821 PIA 2
4 - AY-8910 Sound IC
5 - TMS5220 Speech IC
The PIAs are identical and both socketed so if either of these displays an error than can be swapped for a known good unit or swapped around to see if the fault follows the suspect chip.
I also tried a new set of ROMs just in case as it is difficult to verify ROM 2. ROM 1 (1D) and ROM 3 (1G) can be verified against a ROM image but ROM 2 (1E) cannot as it contains a game's serial number so its image size will vary from board to board. The ROMs made no difference so I proceeded to check the various components.
The board has 3 individual sections with each section responsible for a different type of sound. There is a speech section (TMS5220), sound (AY-8910) and sound DAC (1408).
To test the board properly you either need to run the sound diagnostics with the board in the machine or construct a simple circuit to manually select the sounds. There is a circuit for the test board at
http://www.zaccaria-pinball.com/gen2/index.html. I modified my test switch setup slightly and only set my switches to be open or to connect to ground. The inputs are already tied high so you do not need to worry about floating lines. The sounds are selecting by placing the correct sound number on the input bus of the 74LS244. Bits HS0-HS6 select the desired sound number and toggling bit HS7 triggers an interrupt on the CPU and causes it to read/play the selected sound number.
David Gersic's site has sounds maps for most of the Zaccaria games except for Magic Castle but it does not take long to determine some sample sound numbers for testing. Most Zaccarias skip the first 6 sound numbers (0000 0000 - 0000 0101) and start on 6 (0000 0110). Roughly the first 30-35 sounds are used by either the AY-8190 or 1408 DAC chips and the next 20 sounds are used by the speech. To determine which sound chip is being selected you can watch for data on each of the sound chips inputs. I used pins 27,29-37 on the AY-8910, 8,12-15,17,etc on the TMS5220 inputs and pins 5-12 on DAC 1408 to determine some sample sounds.
I could now select speech, hear music from the AY-8910 (although reduced volume) but nothing from the 1408.
I was able to trace data from the 6821 (1M) to the input of AY-8910, from pins 3,4 of AY-8910 to the MF-20, and into the input of the TDA amplifier. The input only showed about 1V P-P compared to the speech volume at 2V P-P. The schematic showed that R93 which carries the sound into the amplifier should have been a 10K instead of a 56K installed on the board. I installed a 12K resistor in parallel with this to bring the resistance back to the correct value and this rectified the problem with the sound levels. I paralleled the resistor as it can be easily removed if necessary.
The next problem was to solve the DAC 1408 sounds.
After checking the 1408 was being enabled and that it was outputting sound on pin 4, I traced this further through to Q2. This transistor showed a signal on the emitter of the transistor but nothing at the collector. I was unable to measure this in circuit with my meter so after removing this and testing with my meter on diode test, this confirmed it was open. The meter should show a 0.5-0.6 drop across the base to emitter and base to collector pins of the Q2 2N4401. Replacing the transistor solved the sound problem. I then tested all sounds one by one by stepping through all dipswitch combinations.
Occasionally the sounds would not stop working during testing and I found that by lightly pressing on the board around the CN7 connector would solve the problem. I could not see any dry joints around here but decided to touch up the joints anyway. This cured the final problem and got the board fully running again.
