MOTM

I have spent this evening painstakingly taking voltage readings and
calculating the resistor sizes needed for a rotary octave switch in the
MOTM VCO.  Here is information for your disbelief or other general misuse.

Disclaimer:  I am not an electrical design engineer.  I am not suggesting
that you use any of the information I am sending for any reason other than
a comparison to your own.  These are the numbers that worked in my MOTM-300
VCO # 1.  Your milage may vary depending on your exact value of R4.  Here
are my numbers.  I arrived at these figures by beating my oscillators
together across their range and taking actual measurements on the pot side
of R4 to determine what voltage would be required.  As you might expect, I
found a VERY linear scale.  (Thanks AGAIN Paul).

Working from the middle of the 100K pot (zero voltage point), an octave
change in the VCO output is caused by 1.5 volts change on the pot side of
R4. (actual measured normal range 1.497 - 1.503)  So, a two octave shift
would require 3 volts.

Working up the scale, if you wanted a musical 5th you add .875 volts
(actual measured normal range .871 - .882)  So, the fifth above the first
octave up would require 2.375 volts.

Working down the scale from the middle (zero), the musical fifth would be a
different interval, so the negative voltage would be the difference between
the positive octave and positive 5th.  (1.5 minus .875 = .625)  Actual
measured value .620 - .622.

If a series string of resistors was used to create a voltage divider to
replace the VR 5, you could tap the division points with your rotary switch
to get the voltages you need.  I choose to stay with the total of 100K for
my divider (value of VR5).

If you wanted octaves, a 5K resistor would be perfect beween each octave. 
I find only 4.99K in the 0.1% resistors in the Mouser catalog (page 189). 
That's close enough.  Fill up the ends with so you have 50 K total on each
side of zero and a total of 100K.  If you want musical fifths (like I do)
then the ideal resistance would be about 2080 and 2920 ohms instead of the
5K.  The closest listed in Mouser are 2050 and 2940.  They are close enough
according to my calculations.

So, for octaves and 5ths, you would alternate the 2050 and 2940 (an even
number on each side of zero) with the 2940 being the first ABOVE zero
headed to +15 VDC and the 2050 being the first below zero headed toward
-15VDC.  Of course, you need to fill in the ends so you have 50 K total on
each side of zero.

I calculated the resultant voltages from these resistor sizes in a 100K ohm
total voltage divider circuit.  Starting at zero volts on the pot side of
R4, I beat my two VCOs to as near sync as I could get without actual sync. 
I tuned the test oscillator up so that the voltage that I had calculated
for these resistors was present on the pot side of R4.  In every case, the
beat frequency was close enough to the reference oscillator that soft sync
brought the two oscillators together without having to touch the fine
tuning knob.

Sorry to bore you guys that have no interest in octave switches, but I
think several do and this is one easy way to accomplish the task with no
modifications to the board. The combination of resistors and rotary switch
can connect to the same 3 wires that now connect VR5.

Now, for those that might want to speculate on another interesting twist, I
have planned 11 preset positions for my selector switch (octaves and
fifths) corresponding to the 11 tick marks on the panel (0 through 10). 
The 12th position (facing straight down) will be the same as the "wide"
position that Moog put on his Micromoog (and other?)  synths.  In this
position, the fine frequency control will change to a coarse control so
that you still have the larger frequency range of the VCO outside of the
preset octaves on the rotary switch.  A second pole on my rotary switch
will lower the value of R3 in this one position.

OK, that's my take on the easiest way to convert the MOTM 300 coarse
frequency control to a octave selection switch.  I have originally planned
multi-turn trimmers for each position, but Paul has swayed me toward the
0.1% resistors.  Hence, all the effort into measurement and calculation.

Another disclaimer:  None of the above information is endorsed, recommended
for use, or otherwise made legitimate by Paul.  This is all Larry (a
non-engineer) speculating about what he thinks will work OK.  Now, Dave,
John, others, prove me wrong or close to right so I wont feel so bad about
giving this a whirl.

Larry Hendry