Tuesday June 18, 2024

Stern Pinball SB-100 Sound Board

In Theory - How The Board Works
In order to diagnose these boards better, it's best to know how the circuit(s) are actually working.  I wasn't finding much information available on these boards anywhere until I stumbled into a rec.games.pinball post from 2000 where someone took the time to explain their thoughts on how the components on the board are interacting.  This significantly helped me fill in some blanks in my knowledge of how the board is working.  And so I will attempt to bring all of this information together along with my own theories.  If I mistate anything, just contact me and I'll evaluate and fix if necessary.

Theory of Operation

1. The 6800 CPU connects directly to the sound board via a wire harness between the 32-pin connector A4J5 on the MPU board to A6J1 on the sound board.  The sound board is mapped into the MPU board's 6800 CPU, which means there are no independent CPU or PIA chips driving the sound card.  The bad news here is a faulty sound board then has the potential to lock up the MPU board (seems to be pretty common occurrence).

2. Power is supplied to ICs U1 thru U5 through the 32-pin harness between the MPU board and sound board (pin 30 on J1).  The rest of the ICs on the board are powered from the 11.5v that enters the sound board at connector J2 (pins 5 & 6).  The 11.5v is broken into 11.5v, 10v, 6.2v and 5v power circuits (though on the Rev-3 boards without the CHIME circuit there would be no 6.2v power circuit.

Power

3. During boot-up sound is disabled until the CPU's /RESET goes HIGH.  There is a reset circuit consisting of the /RESET line from the 6800 CPU on the MPU board, a 74107 (U6) "Dual J-K Flip Flop", 2n3904 NPN transistor (Q6) and a few resistors.  When the /RESET line goes low, the chip is cleared and /Q (U6 pin 6) is HIGH which activates the Q6 transistor.  When Q6 is activated, it would pull the bypass pin 1 on both LM380 amplifier chips (U20, U21) to ground.  Normally pin 1 on both LM380 chips is "bypassed" via a 10uf capacitor (c42) and this would reduce power supply noise on the LM380's output.  Instead, when Q6 activates, pin 1 on the LM380 is pulled to ground and the LM380 is muted (*note: other sources seem to indicate muting would occur when pin 1 is pulled to vcc, so more research here may be needed).

Reset Circuit
Picture: Reset Circuit

 
4. U1 (7402 NOR-Gate) and U2 (7411 AND-Gate) outputs are used to enable the 7475 "4-bit Bistable Latches" ICs (U3 thru U5) at the CLK pins.  The inputs to these latches are the data lines from the 6800 CPU.  Depending on the sound being played, any one of the data lines D0 thru D7 could be pulsed.  The 7475 would read the input and set its output to a 0 or 1.  The latch will remain in that state until the input is pulsed again by the CPU.

If the SB-100 has U5 (for chimes) then the DIP switches on the MPU board that are read during MPU bootup (and stored in RAM) would select between chime or electronic sounds.  U2 would either enable both U3 and U4 for electronic sounds, or ONLY enable U5 if chime sound was selected.

U1 thru U5
Picture: ICs U1 thru U5

5. The outputs on U3 and U4 7575 ICs are connected to one of the input pins on U7 (a 7408 "Quad 2-input AND Gates").  As the "Quad" in the name suggests, the U7 IC has a total of four "2-input AND Gates".  Each of these AND gates has two inputs.  The first input as mentioned previously is coming from the 7475 bi-stable latch.  The second input is the frequency being created by a multivibrator circuit.  There are test points TP2, TP3, TP5 which can be used to test the frequency output of the multivibrators.

U7 7408
Picture: U7 pin 5 enables the frequency on pin 4 to pass to the output pin 6

6. Multivibrator circuits are used to create 3 distinct sound frequencies.  These are shaded in red below.  They each consist of an LM324, potentiometer, a capacitor, and a few resistors.  They are always running and creating a frequency at their outputs.  The frequency can be adjusted via the potentiometer in each of these circuits and according to an RGP post are adjustable between 1-3.3kHz, although after they are fed to the 4013 ICs it would be halved (and once again at TP3 since the frequency goes through yet another 4013).  These frequencies are used for both the electronic sounds and chime sounds.

Multivibrators
Picture: Three separate multivibrators using LM324 (U17), potentiometers, resistors, capacitor

6. For electronic sounds, the frequency output of the multivibrators are piped into the data input (D) on a 4013 "CMOS Dual D Flip Flop".  Here the logic state at the data input (D) on the 4013 is transferred to the (Q) output on the rising edge of the clock signal / frequency.  In doing this, the 4013 acts as a frequency divider in which the output (Q) pin 1 is half the input frequency (CLK) pin 3.  An oscilloscope hooked up to TP2, TP3 or TP5 will show a square wave.

U7 7408
Picture: Output at 4013 pin 1 would be half the frequency of the multivibrator


Square Wave at TP2
Picture: Square Wave at TP2 on a Stern Hot Hand SB-100 Sound Board (Left); Frequency Reading on VMM (Right)

The 4013's creates the second input on the 7408 IC.  The frequency is always at the 7408 IC's input pin, waiting to be enabled by the other input to the 7408 Gate (again coming from the 7475 IC).  When both inputs are HIGH, the output is also HIGH and the frequency is allowed to output on the 7408's output pin for that Gate.  This signal would then feed to the final output stage.

7. For chime sounds (on boards with a U5 IC), the frequency follows a slightly different path.  The raw frequency / pulse generated by the multivibrators appear to be overlayed on the square waveforms.  In the picture below, the blue arrows show where the multivibrator output is tapped directly & the red arrows show where the 4013 half-frequency output is tapped.

Multivibrators 2
Picture: Blue arrows are frequency from multivibrator; Red arrows are half-frequency from 4013


Both of these frequencies for each tone are used as an input to the LM324 ICs on the bottom-right of the schematic.  The output of the LM324 is used as the input to the MC3340P electronic attenuator ICs.  I'm not 100% sure but I would guess that the resulting sound waveform at MC3340P's pin 1 would be triangular or sawtooth instead of square.

Overlayed
Picture: Two waveforms being used for the non-inverting input pin 3 (+) to the LM324

According to its Motorola Datasheet, the MC3340P is primarily used in volume control or compression/expansion amplifier applications.  Each of the chime circuits has its own MC3340P.  I found this interesting at first until I thought about it more.  There is only one other MC3340P used on the SB-100 and that's at U14 that has the cabinet volume control potentiometer hooked into it on pin 2.

Taking it back to basics lets think how physical chime units work in older pinball machines.  A solenoid would be activated causing a plunger to wack a chime.  The initial impact would sound the chime VERY loud and resonate at its correct frequency and slowly fade in volume.  So a separate MC3340P used in each of the chime circuits makes a lot of sense if it's used to create that "fade sound into the distance" effect.  So that's my theory until someone tells me otherwise!

Let's look at the other half of the chime circuit feeds into pin 2 on one of the MC3340P ICs.  Going all the way back to U5 (7475) which again would be switched on at the DIP switches on the MPU board, we see the following circuit for the output on 1Q.

U5 Chimes
Picture: Other half of chime circuit

There's a common 2n4403 transistor, some resistors and capacitors and an LM324.  The output of the LM324 pin 1 then feeds to MC3340P pin 2 (the pin that could have a potentiometer output sent to it to control the volume).  So, the circuit above seems to be entirely chime volume related and what creates that "fade" effect.  Picture manually throwing a volume knob HIGH and then slowly turning it LOWER until the sound is no longer heard.

How do I think this happens?  The CPU would tell the 7475 to play the chime sound through pin 2 (1D) and this would cause the output (1Q) to latch at 5v.  This would then cause the 2n4403 (Q5) transistor to activate and the capacitor C21 would dump its charge to the path of least resistance (r47 in circuit above) to pin 2 on the LM324.  The output of the LM324 would be a voltage frequency akin to adjusting a potentiometer HIGH quickly, then LOWER as the capacitor is drained further and further.  During all of this, the CPU would need to latch the 1Q input back to 0v to stop the chime from sounding.  This is just my theory on what's in the schematic.  If I'm wrong please contact me and I will update this info if necessary.

So with that, we have the sound waveform input on pin 1 of the MC3340P and the volume input on pin 2.  These then go to the final output stage.

8. The last thing to discuss as far as sound generation are the outputs on U4.  These outputs appear to be used only when electronic sounds are selected for the game. 

Noise Generator / Triangle Wave
Picture: Noise Generator / Triangle Wave Generator

The 3Q output looks like it would generate a triangle waveform based on the two LM324s circuits used.  And interestingly enough, the first part of the circuit with the 2n4403, capacitor and resistors resembles the chime circuit.  So a varying voltage is being created using Q1, the 3.3uf cap and resistors and that's fed into the LM324.  This is then used to create a frequency for the LM324 inputs.

The 4Q output looks like a noise generator and uses a 556 dual timer IC and 74107 "Dual J-K Flip Flop".  When 4Q is latched to 5v it would reset the 556 timer and start producing noise based on the configuration of the trigger/threshold/discharge pins.  The output is sent to the 74107 (U6) pin 12 as the CLK.  This would create a frequency at 74107 pin 13.. I would again guess here that it would be a square wave.

That's it for all the sounds on the board, lets look at the final output stages.

9. All of these sounds, regardless of where they're created end up at the same output stages consisting of yet another LM324 (U18), an MC3340P (u14) used with the volume pot, and two LM380 ICs (U20, U21).

First the sound frequencies enter an LM324 (U18) at pin10.  This LM324 acts as a voltage follower and gives effective isolation of the output from the source signal.  Very little power is drawn from the signal source so you avoid "loading" effects.  In any case, voltage at pin 10 on U18 should be the same as on pin 8.

Output 1
Picture: LM324 Voltage Follower & MC3340P Electronic Attenuator

The output of the LM324 is sent to the MC3340P (Electronic Attenuator) input pin 1.  Here the MC3340P acts at the volume control for the entire machine.  The potentiometer seen in the schematic above is the volume pot inside the cabinet.  As you adjust the volume, the MC3340P does basically the opposite of an amplifier and reduces the volume.


Output 2
Picture: LM380 Power Amplifiers used in "bridged" configuration

The output of the MC3340P on pin 7 then feeds to a voltage divider consisting of r19, r72 and finally makes its way to the LM380 power amps.  The power amps are used in a "bridged configuration" which doubles their output from around 2 watts to 4 watts.  The output of these power amps is then sent through to the speakers.

That's it!  That's the basic theory of how the sound board works from everything I have put together so far.

So how do you diagnose these boards?  Continue onward..

page1 page2 page3 page4


Comments:
Stern Nugent Sound Board needed
Posted 05/03/12 1:53PM by nowears
Please let me know when / if you have a stern sb-100 sound board available. My Nugent does not have one.
Thanks,
Re: Anonymous Techdoser
Posted 12/31/11 2:41PM by AceBHound
Hi Bob, contact me via email on the "Contact Us" page on the left-hand navigation menu on this website. We can discuss details further.
Anonymous Techdoser
Posted 12/26/11 4:04PM by Anonymous Techdoser
Hi, Do you have an SB-100 for sale and how much? Would you give a discount for my non-working board? If not, how much to repair mine? Thanks, Bob
Re: Great Info
Posted 12/05/11 6:38PM by AceBHound
I may have an extra SB-100 available soon -- I'll email you. Also for anyone else interested, I will likely have more for sale down the road & entertain repairing some boards for people as time permits if they aren't severely hacked. Repair costs would depend on how "dead" your board is. Email is on my Contact page on this website.
Great Info
Posted 11/03/11 4:43PM by bobcav
This is great info. I'm currently looking for an SB 100 sound board as my Nugent machine has never had a board since I've owned it. Any thoughts on how to acquire one, either in working order or not?
Stern SB 100 sound board
Posted 03/10/11 8:24AM by Anonymous Techdoser
what a great help!
The test point voltages certainly gave me the confidence to make a repair to a Stern "Magic".
I had replaced the chips in their entirety, capacitors, transistors, zenner diode and a couple of blown resistors. Until I found your voltage listing I could not have any confidence in the work I had done. I soon traced the problem to a cracked lead on the circuit board, in the voltage splitter prior to the amp chips.
Thanks!!!!