Trimming Issues

[thismatters]

Hello, I've been working on a clone of your wonderful module for my DIY synth. I'm in the breadboard phase and was having some difficulty trimming the randomness to the suitable range. It seems I'm not the only one!

I'm using a 25 turn trimmer on my breadboard, so I'm really able to hone in on a perfect setting so that the loop controlled noise (pin 6 & 13 of the 4016) tracks smoothly with the change knob. Admittedly I'm using a vanishingly small sample of transistors for my experimentation, but what I'm seeing is that the trimmer resistance is less than 200K Ohms. Even with the 25 turn trimmer finding the sweet spot is troublesome when using a 1M Ohm trimmer; doing so in a single turn would require the dexterity of a safecracker.

To anyone having troubles trimming your Turing Machine, I recommend trying a 200K Ohm or 500K Ohm trimmer of the same footprint.

I've also found that a different transistor choice can improve the trimmability of the module. The BC547C appears to have a higher baseline noisieness which requires less amplification achieve the necessary range (Note: the BC547C has opposite pinout as the 2N3904). Swapping to this transistor brings the necessary trimmer resistance to 5K Ohm, affording a much wider range of acceptable calibration when trimming.

Again, I haven't studied this systematically to see if variations in batches of transistors makes the full 1M Ohm range necessary for such a widely adopted and loved module, but given the difficulties that so many have reported in trimming the module it seems important to raise this issue.

Thank you.

Below is a list of the issues that seem to have difficulty in calibration as a primary factor.

Likely trimmer value issues:
#57, #61, #97, #109, #141, #156, #168, #174, #196, #200

Possible trimmer value issues:
#52, #149

Not a lot of conversation, but plausibly trimmer related:
#82, #93, #161, #180