Oakley midiDAC User Manual - page 8
Each output section is called a ‘sample and hold’, although to be strict the demultiplexer also
forms part of the sample and hold. The capacitor in each S/H holds or stores the voltage that
is briefly connected to it. The op-amp that is connected to it, allows this voltage to be ‘sniffed’
without effecting the actual value. The op-amps are connected as voltage followers or buffers.
They have gain of 1. Thus, the sampled voltage on the hold capacitors can be found at the
output of each op-amp. Note, that pitch CV and pitch bend use low offset FET op-amps, U12,
for accurate pitch control.
Note that not all of the eight outputs from the demultiplexor are connected to S/H circuits.
The PIC only processes six midi controllers so the other two outputs are unused.
The pitch and pitch bend CV are processed further by the midiDAC. This circuitry is seen on
page two of the schematics.
The pitch CV is sent to the slide circuit. This circuit is based heavily on the slide circuit from
the TB303, and, of course, the Oakley 3031. When the slide is not enabled, the first portion of
the analogue switch, U5, is off. The pitch CV is then passed through the slide pot straight to
the op-amp buffer, U4 (pins 1, 2, 3). The resistance of the slide pot has no effect on the CV
because the input impedance of the buffer is very, very high. The second portion of U5 is on,
and the capacitor, C2 is charged up to the CV voltage. Slide is activated, either by the PIC via
the SLD logic line, or manually via the ‘SLID’ header being shorted by the SLIDE switch.
This then causes the two sections of U5 swap states. The pitch CV now has to charge C2, via
the slide pot, every time the CV changes. The higher the resistance of the slide pot, the longer
it takes to charge up or down.
The slide signal from the PIC also drives the ‘legato’ LED via U6. The LED is connected to a
2 pin header on the PCB, labelled ‘LEG’.
The pitch bend CV varies from 0 to 10V depending on the status of the pitch bend controller.
For normal use, we require the pitch bend to go from -5V to +5V, with 0V representing the
pitch bend wheel centralised. To do this we must subtract exactly 5V from our pitch bend CV
signal.
This is done with a simple summing amplifier based around U2 (pins 5, 6, 7). This adds the
-5V reference voltage to the pitch CV. Since the pitch CV is centralised at +5V, when we add
these two voltages together, they cancel. However, a positive (upwards) bend produces a
negative voltage so we must invert the summed output with another op-amp circuit. This is
based around U4 (pins 5, 6,7) and features capacitive loading protection via R28 and C8. The
output of this circuit goes to the Pitch bend output socket and the Pitch Bend depth pot.
The Pitch bend depth pot allows a fraction of the pitch bend CV to be added to the pitch CV.
Thus wiggling the pitch bender will automatically control the pitch of any connected VCOs.
The circuitry based around both halves of U1 add the pitch bend CV to the pitch CV and to
allow fine tuning of the VCO pitch.
Note the use of 0.1% resistors in the summing circuit. If ordinary 5% resistors were used here,
it is likely that the -5V and the +5V signals would not be exactly cancelled. This would result
in a small error voltage, ie. non zero, at the pitch bend output.
8