[sdiy] DOTCOM Analog Sequencer.. next project startup..

[Brian Willoughby]

On Jul 25, 2020, at 3:54 AM, Ingo Debus <igg.debus at gmail.com> wrote:
>> Am 25.07.2020 um 09:20 schrieb rsdio at audiobanshee.com:
>>> I once had the idea that for an user interface like this, you could use *one* rotary encoder that is driven by many thumbwheel-style knobs, all one the same shaft. When one of the thumbwheels is touched, the corresponding parameter is selected. How about that? Of course you can only change one parameter at a time, so this is no good for mixing consoles or the like.  But for a sequencer this might work.
>> 
>> That's a neat idea, but unless someone has already manufactured the complicated mechanical structure that would allow 16 thumbwheel to turn the same encoder (without turning the other thumbwheels), then it sounds like a very expensive contraption to design and put together.
> 
> Well, I thought all thumbwheels would be turning simultaneously. Would that be a problem? You turn one and the others follow.

I imagine that the problem would be friction. Thumbwheels are always detented, with a reasonable amount of friction to keep them on the selected value until enough pressure turns them to another value. With 32 or 16 in parallel, the friction would be multiplied

> The tricky part is detecting which one is touched.

Indeed. It would surely require wires for each separate sensor, and those wires would get tangled up as the one encoder is rotated. Any sort of wireless solution would probably have to be mechanical.

> Ingo

Thinking about your idea, perhaps each thumbwheel could rotate on its own axle, but that axle would be captured by a slot. Spring-loaded momentary buttons on either side would push the shaft towards the face plate, but still allow rotation. Pressing on an individual thumbwheel with enough force to rotate it would also press the axle against the tack switches (at least one of them, if not both), and that would indicate to the circuit which one is being touched. Meanwhile, the inner edge of the thumbwheel would be pressed against a long shaft behind, and only the thumbwheel being rotated would be in contact with the shaft. Either the thumbwheel surface or the shaft would need to be rubberized or somehow produce enough contact friction for the thumbwheel to turn the shaft. Most thumbwheel s have a sharp edge with ridges for fingertips to catch, but it could just be a rubber surface instead of a hard surface. This mechanical arrangement would actually allow each thumbwheel to turn on its own, but the switches would tell the circuit which value should be adjusted by the common shaft.

The challenge might be to keep the virtual value aligned with any numbers printed on the face of the thumbwheel if the user doesn't press hard enough to rotate the common shaft. The solution here is probably to not have any markings on the thumbwheels. The advantage is that the value could have a much greater range than 0 to 9. Also, just like a standard encoder, if things get out of sync, the changes are still relative, so it's easy to adjust.

All said and done, it would probably still cost more than 16 or 32 rotary encoders, but it sure would present a unique user interface.

Brian Willoughby