Polysix - Sharing my Arduino Code

As many of you know, I made some pretty substantial modifications to my Korg Polysix.  One of the biggest changes was to replace the keybed to enable aftertouch.  To make this happen, I had to replace the Polysix's Key Assigner with an Arduino.  It was a lot of work, but once I made the core hardware modifications, I found that I was able to add lots of features to the Polysix, such as: aftertouch and portamento, more interesting detuning for unison mode, a poly-unison mode, a Moog-like single-trigger mode, and a sustain pedal.  All of these modifications were effected through software changes on the Arduino.

Since I have made those modifications, I have had some people ask if I could share the software.  At first I declined, because I was embarrassed of the poor quality of the code.  I also didn't really know how to share it.  Then, I learned how to use GitHub.  So, I at least know how to share my code now.  All I have to do is get over my embarrassment.  Well, here we go...

If you want to look at some nasty code for how to replace the Polysix Key Assigner using an Arduino, you can find my GitHub repository at: https://github.com/chipaudette/ArduinoPolySix.

Yes, my code is tortured and very difficult to follow.  But, that's what happens when you're just learning how to do this stuff.  Sorry.  If I were to re-write it from scratch it would look a lot different...but this is what I have now.  Maybe you'll find something useful.

Thanks for reading!

Portastudio - Voltage Control of Speed

After my experiments with live looping using my Portastudio 424, I keep thinking about what interesting modifications I can do to push the Portastudio into unexpected places as a performance instrument.  One idea was to expand the user control over the playback speed of the tape so that I could arbitrarily vary the pitch of the recorded audio.  Sure, the 424 gives some control over the tape speed, but it does not allow you to smoothly span a wide range.  So, I started looking into how I could modify the 424 to give me what I wanted.  Ideally, if I truly get aribrary control over the tape speed, I could turn my Portastudio into a monophonic Mellotron, which could be kinda cool!  Here's the story of what I've discovered....

Built-In Tape Speed / Pitch Controls on my Portastudio 424

Built-In Pitch Controls:  As shown in the picture above, the Portastudio 424 comes with two user controls over tape speed.  First, there is a 3 position switch that sets the gross tape speed to either "slow" (15/16 inch per second), "normal" (standard 1 7/8 inch/sec), and "high" (3 3/4 inch/sec).  Assuming that it is usually set in the "normal" position, this switch lets you either drop the pitch by an octave or raise the pitch by an octave.  If you want finer control, you can use the pitch adjust knob.  Unfortunately, this only allows you to lower or raise the pitch by about 2 semitones.  So, there is a big gap between the gross tape speed switch and the fine pitch adjust knob.  I would like to modify the Portastudio to allow smooth changes between any pitch.  Let the hacking begin!

Opening the Portastudio, you can see the cassette and motor
assembly in the upper right.

The Cassette and Motor Assembly:  My first step was to open up the Portastudio to see how the cassette playback assembly is put together.  As you can see above, opening the Portastudio reveals a big circuit board that surrounds an assembly that holds the cassette, motors, and other mechanical bits.  Looking in detail at the cassette and motor assembly (picture below), you can clearly see the thin metal capstan and its corresponding pinch roller.  The capstan (and its roller) are what pull the tape over the playback head.  It is the capstan that controls the speed of the playback.  Controlling the speed of the capstan is the key to controlling the speed of the tape playback.

Zooming in on the cassette and motor assembly, you can see the capstan
and its mating pinch roller.  This is what controls the tape speed.

Capstan Motor:  Removing the cassette and motor assembly from the Portastudio, the picture below shows its underside.  It is a very complicated mechanical assembly with multiple motors, gears, belts, and circuit boards.  Staying focused on just the elements associated with the capstan, one can see a big flywheel that is directly attached to the backside of the capstan itself.  The flywheel is driven from a belt attached to a the capstan motor.  The capstan motor is attached to a circuit board via four electrical terminals.  This must be how the speed of the capstan is controlled.

Looking at the underside of the cassette and motor assembly.  You can
see the capstan flywheel, which is belt-driven by the capstan motor,
which I'm probing using my clip leads.

Four-Wire Connection:  Frankly, I was surprised to see that the motor had four wires to control it.  Not knowing anything about cassette players and capstan motors, I expected to find a basic DC motor with just two terminals.  I expected that the speed would be controlled by changing the DC voltage of the positive terminate or by using PWM of the the positive voltage applied to the motor.  Apparently, that's not the case.  How is this motor controlled?

Rotating the cassette and motor assembly again.  Here you can see
that the capstan motor terminals are labeled 1-4.

Looking at the Schematic: When it isn't clear how something works, it is usually best to consult the schematic.  Sadly, I don't have a schematic for the Portastudio 424.  Even scouring the internet (and visiting some unsavory web forums) did not yield the schematic.  The closest that I got was on eBay where someone was selling a paper copy of the Service Manual (which includes the schematics) for the Portastudio 464.  Since the 464 and the 424 were out a similar times, my hope was that the schematics would be close enough to enable my hacks.  So, I ordered it.  When I arrived, I found that the capstan motor and speed controls (in the 464) were wired as shown below.  It clearly shows the four wire connection -- two connections for power (12V and ground) and two connections ("B" and "A") that loop out through the user controls.  These must control the speed.  But why are the connections configured as a loop, instead of just injecting a commanding voltage signal from the outside?

Schematic from the Portastudio 464 Service Manual.  Hopefully, it is similar to my Portastudio 424.
Probing my capstan motor, it appears that the motor terminals are reversed...on my Portastudio 424,
Pin 4 is ground, Pin 3 is +12V, and Pin 2 and 1 must be the ones shown above as "B" and "A".

Capstan Speed Control:  A little research on the internet yielded some information.  The key idea is that the capstan must spin at a very precise speed (or else you get "wow and flutter"), so capstan motors are built with some sort of closed-loop control system.  This means that they have a method of sensing speed and then adjusting their speed based on the sensed value.  In many cassette systems, this closed loop feedback is entirely within the motor body.  In more flexible systems (such as the Portastudios), two sense and command signals are brought out to allow for user adjustment of the pitch.  In this case, the sense and command signals must be terminals "A" and "B".  I do not know which is which.  So, I'll probe the signals to see what I get.

Assuming Motor Pin 4 is Ground, Here are the Voltages that  I Measured
 at Different Tape Speeds.  Pin 1 Appears to Control the Tape Speed.

Finding the Command Terminal:  Using my digital multi-meter (DMM) with the Portastudio turned on, I quickly determined that on my 424, Pin #4 is ground.  This is in contrast to the 464 schematic, which showed ground as Pin #1.  Regardless, moving forward, I put in a cassette and pressed play.  Then, using Pin #4 as a reference, I measured the voltage at the other three terminals for different speed settings.  The values are shown in the graph above.  As can be seen, the voltage at Pin #1 drops as the speed increases.  The voltage at the other pins stays the same (basically).  So, I think that this confirms that Pin #1 is the terminal for the speed command signal, because it is the only one that changes with speed.  It appears likely that Pin #3 is the power supply for the motor, because it is the highest voltage signal and stays firm at +12V.  Therefore, through process of elimination, Pin #2 must be the terminal associated with the speed sensing.

After recording a ~65 Hz tone at the normal tape speed, I see that there is a nice
inverse relationship between the motor control voltage and the resulting pitch.

Command Voltage:  According to the 464 schematic, the speed command voltage is set by drawing current out of the speed sense pin (Pin #2 in the 424), and altering the voltage delivered to the command pin (Pin #1 in the 424).  Therefore, we would expect the voltage at Pin #1 to be lower (ie, negative) relative to Pin #2.  The blue line in the graph above shows voltage measurements that I made that confirm this expectation.  As you can see, the slowest speed exhibits a voltage at Pin 1 that is about -0.5V relative to Pin #2.  As the speed is increased. the voltage at Pin 1 eventually reaches about -3.5V relative to Pin 2.  Again, this is the relative voltage...the absolute voltage (well, relative to ground) was shown in an earlier graph and is running between +8.5V and +11.5V.

Pitch vs Speed:  So now that we have a good handle on how the Portastudio changes the command voltage based on the user settings for the tape speed, there is still the question of how much the speed of the tape actually changes.  Since tape speed directly affects the pitch of the audio that has been recorded to the tape, I've decided to quantify the change in tape speed by recording a tone onto the tape (at "Normal" speed) and then measuring the frequency of the tone as I change the speed.  The results of this experiment are shown as the red line in the graph above.  As expected, higher tape speed results in higher frequency.  Plotting frequency directly against the capstan command voltage, we get the relationship shown below.  Nice and linear.  Excellent.

The relationship between control voltage and pitch is nice and linear.

Creating the Command Voltage:  OK, now we're getting serious.  If I really want to make a mellotron out of my Portastudio, I have to generate the precise command voltage in order to achieve the pitch that I want.  The "command voltage" is is the voltage at Pin #1 relative to Pin #2, both of which are riding well above ground.  Assuming that I'm creating my command voltage from a 0-5V sounce (such as from an Arduino or something), I need a circuit that will both boost the voltage up to the correct DC level and to adjust the voltage based on whatever is happening on Pin #2.  The circuit shown below should do exactly that.

Notional Op-Amp Circuit Configuration to Generate the Correct
Capstan Command Voltage (Pin #1) Given a 0-5V Input Voltage.
[Revised, Mar 9, 2014]

Picking the Right Input Voltage:  While the circuit above should achieve my goals, it is an inverting amplifier with respect to the input voltage that I'll be injecting.  This means that an upward-going input signal results in a downward-going output signal.  As a result, the relationship shown earlier in the frequency-vs-voltage plot shown earlier is inverted.  To account for the inversion, I made the graph below.  Also, because I actually care about the perceived pitch of the tone (the note on the scale), not the actual frequency of the tone, I converted "frequency" axis into "pitch".  With this new graph, I can choose how many semitones of pitch shifting I want (say, shift up an octave, which is 12 semitones) and then I can simply read off the voltage that I need to send to my op-amp circuit (in this case, +12 semitones requires a voltage of +3.25V.  Easy!

If I want to achieve a certain pitch, here is the relationship that
determines the control voltage that I need to generate.

Next Steps:  With all of this analysis, I think that I have figured out how to control the tape speed of the Portastudio 424 to get the pitch shifting that I want.  Now I have to go and try it out.  It's time to smell the solder!

UPDATE, Mar 9, 2014:  I tried building the original op-amp circuit that I showed.  It did not have the correct input-output relationship at all.  The problem was that I did not include any of the resistors (R3 and R4) that are now shown on the non-inverting input.  Ooops!  Adding R3 and R4, the input-output relationship is now correct.