> Some observations:
>
> A-149-1 Quantized/Stored Random Voltages:
> - N+1 output never "jumps" - it repeats the previous value or changes
> to next lower of higher value.
> - 2^n output never gives same value twice in succession - expect for
> the lowest and and highest possible values, which it repeats quite
> often.
> - "Equal probability distribution" output has some "personality" too.
> - "Adjustable probability" distribution output makes distinctive
> patterns: For example, if D=0, then the output typically jumps up and
> then settles down in succesive steps. If D=10, this is mirrored.
>
> A-149-2 Digital Random Voltages module:
> A-149-1 knobs don't affect A-149-2. Outputs 1 to 7 are "echoes" of
> output 8. In this order:
> 8 => 7 => 5 => 4 => 2 => 1 => 3 => 6
> For example, output 5 is two clock cycles behind output 8.
>
> (I assume, I would prefer more "true" randomness.)
>
> Best regards
> Jari Jokinen

The A-149-1 just simulates the quantized/stored random voltages of the
Buchla module, i.e. the outputs of a shift register with two or-ed feedback
lines are weighted (with different resistor values in the Buchla circuit, by
calculation and DAC in the A-149-1) to obtain a random voltage. The circuit
is similar to a digital noise (which is also not really random but has a
period depending upon the clock and the shift register length). Due to the
nature of this circuit the randomness may be different from what you expect.

The behaviour of the A-149-2 is normal as it's nothing but the shift
register outputs. Please look at the A-149-2 description " ... correspond to
the 8 digital outputs of the shift register that is used to generate the
Quantized Random Voltages". Consequently the shifting of the states from
output to output is the normal behaviour.

Best wishes
Dieter Doepfer