Blue Box delay in octave shift?

[carboncomp]

I'm building a Blue Box but having a small problem with the octave shift.

When playing a note the fuzz is there, but it take a half second for the signal to shift 2 octave down.

What parts should I be switching out to try and improve this?

[Mark Hammer]

You could only "fix" this by delaying the rest of the signal so that it starts at the same time as the sub-octave.

Consider that a fundamental has to be "counted" twice for a sub-octave to be produced by the flip-flop.  So, if I was to play a note at middle A (440hz), that note would have to be counted twice to start being produced.  That's a note two octaves down, or 110hz, or just under 10msec delay.  Play a lower note and the delay is even longer.

Digital pitch shifters and harmonizers used to have even more delay, but have sped up in recent years.

[carboncomp]

You could only "fix" this by delaying the rest of the signal so that it starts at the same time as the sub-octave.

Consider that a fundamental has to be "counted" twice for a sub-octave to be produced by the flip-flop.  So, if I was to play a note at middle A (440hz), that note would have to be counted twice to start being produced.  That's a note two octaves down, or 110hz, or just under 10msec delay.  Play a lower note and the delay is even longer.

Digital pitch shifters and harmonizers used to have even more delay, but have sped up in recent years.

[Seljer]

The bluebox is quite temporamental about how you play it. Mine sometimes seems to track the 1st harmonic (1 octave up) instead of a the fundamental, giving me only one octave down alltogether on certain ranges. So the standard approach applys: neck pickup, tone rolled off, play round the middle of the fretboard and pick with your fingers instead of using a plectrum

[Mark Hammer]

Well, first, do you know it is a half second, or are you just guessing because there is a detectable delay?  As I indicated, there should be a brief delay between the fuzz and the appearance of the sub-octave in a properly functioning Blue Box, because the flip-flop has to detect two successive peaks before producing an output.  If the delay is clearly, and reliably, longer than that, then there is a different sort of problem and the circuit is not functioning properly.

Alternatively, as Seljer proposes, it may be a case of the signal you are feeding it being sufficient to produce a fuzz, but insufficient to trigger a sub-octave right away for some reason.

[Jazznoise]

These are the joys of a Blue Box. Feeding it a fuzz or wah-d circuit and listening to it bounce along the harmonics is a whole experience in itself. Even open strings interact strangley with it. There are much more stable octave down circuits, but this one is highly idiosyncratic for a pitch shifter.

[Mark Hammer]

I wonder if there would be any percentage in enhancing the gating properties of the unit by sticking a back-to-back pair of silicon diodes in series with that 150k input resistor.

The two transistors in the mixing section are gated by the envelope follower subcircuit.  That certainly helps with removing sputter and other noise.  But it bears noting that they only gate the output of the octave division and fuss.  One thing that might help it behave better is to apply some gating on the input, so that only signals good enough to elicit dependable triggering of the flip-flop make it through.  Insisting that the input signal always be greater than the forward voltage of some diodes to pass and be amplified might help in that regard.

I'm gonna try it this weekend and see if it makes the expected difference.