[sdiy] optical retrofit for minis?

[Kevin Lightner]

Lest I never complete a project that I desperately need to have done, 
I'd like to share what I've been spending a fair amount of time on 
lately. Any practical help would be hugely appreciated.

Here's the situation:

Minimoog keyboard actions are dying left and right.
More and more frequently, the plating on the contact springs and 
bussbars are wearing through, springs stretched, buss bars bent and 
the rubber standoffs on the contact board hardening. I have several 
minis here, all with restored electronics, but with bad keyboards.
The clients are looking into options (buying another mini for parts, 
MIDI, etc.)

I want to offer both my direct clients and as saleable product to 
anyone else (discounted to service centers), a drop-in board with 
photo interruptors. The idea is sound and possible to do.
The SCI T8 and Synclavier did this also.

So, I ordered some various interruptor devices and found one that 
meets all my needs in terms of price, availability, current draw, etc.
Then I made a small prototype from an old Mini action and tested it.
By using the original white, L shaped contact actuators or "pullers" 
as a shutter, the sensors worked great. No additional goodies would 
need to be added or even these white pieces requiring blackening.
During the experiments, I also found that an incredibly controllable 
aftertouch is also possible, similar to the scheme used in some 
Polyfusion keyboards.
While the optosensors I used specified a much higher current draw for 
their IR led, I found that it would still reliably work *much* lower 
too. So, 44 could conceivably be powered at the same time, if 
scanning wasn't implemented.

A PIC could be used for this project and if time was spent on some 
additional programming, I figure the following features could also be 
added:

1. Transposition
2. Inversion
3. Various scales
4. High, low or last note priority
5. Midi in and out (out being rather rare as a feature on minis)
6. Arpeggiator (lots of work, but possible)
7. Velocity and aftertouch (possible, but likely not easy to implement.)
8. Glissando
9. Single/multiple triggering

Several other possibilites too, but these are off the top of my head.
Even an autotune could be made if the thing was taken to the nth degree.

Ideally, a PIC would either have enough I/O lines for every key 
directly, or have enough lines for decent mult/demult schemes with 
additional ICs, one or two 16 bit DACs, a timer or two and draw very 
little current. I've found such a processor, but have no development 
tools.
A fast clock rate and a timer would aid in providing dynamics, which 
*can* be done with just one optosensor.
I understand machine code programming better than I understand a 
windows interface though ;-)
Programming one here would be an expensive ordeal and take considerable time.

Programming of various modes for the user would be via holding down 
the topmost and bottommost keys for 2 or 3 seconds, then making a 
selection via another key. Sort of how the OSCar does theirs.
This, so no additional controls are added.

Another solution would be installation of a Fatar action.
They're cheap, but I feel many people wouldn't want them in their 
model D minis.
Don Martin of Moog Music Cincinatti used this and a small board for 
scanning and D/A, but I never documented his work close enough 
regarding this board and don't have a Donimoog to copy.
He had a 20 pin IC in his though, so I'm assuming he had a small PIC 
in there himself.

Other solutions are welcomed, though if a laser is suggested, a 
solution (two lasers?) would be needed for note priority. A modulated 
laser could identify what note was pressed by pulse counting/timing, 
but once a shutter (key pressed) is made, the beam would be broken 
and note priority would be lost.

So here's the deal: I need some help. Programmer, engineer, board 
person, financer.
I could do it all, but it would arrive *very* slowly.
If any folks on the list feel they would like to contribute to the 
project, I'd be most happy to hear from them. Unfortunately, all work 
would be on spec. That is, payable after the project turns a profit.
Or I could trade them parts, etc, depending on what I have.

With roughly 13,000 minis made, and obviously many that gave their 
lives, visited Studio Electronics, sitting in closets or whatever, 
I'm guessing there may be 100-600 potential customers for the product.
Priced correctly, the product could still turn a tidy profit even on 
these small numbers though.
The action would never need cleaning, would feel better (it doesn't 
have to pull contact springs any longer) and would offer several 
attractive new features.
I have a ton of customers right off the bat.
Two models could even be offered. A bare bones "just the notes" model 
and a "bells and whistles" model with all features. A upgrade rom or 
USB interface could turn the former to the latter, as well.

Any thoughts? Advice? Contributions?
Keep in mind, I've been thinking about this concept for a long time 
and already have done considerable footwork. The concept works. I 
just don't have the resources. :(

Finally, a passive, switch only board for Taurus 1's needs to be done.
I may do this myself, but if anyone would like to chime in on this 
project feel free.
In case you're not aware, the T1 has a huge, inherent design flaw: 
every switch contact must be normally closed or any pedals above an 
open contact won't play. Since the switch contacts are silver and 
tarnish easily, this is an often reoccuring problem where the higher 
you go on the pedal board, the more chances of failure are presented. 
A single, long board (thicker than what Moog used) and long-life 
microswitches would do the job. Unfortunately, I don't have the 
equipment to even prototype such a long pcb. Any ideas on this would 
be welcome too.

Thanks for listening and possible suggestions, etc.
Investors are hugely welcomed also ;-)


-- 
Regards,
Kevin Lightner

Myself: http://www.synthfool.com
Service and sales: http://www.moogmusic.com/service.php
Sales: http://www.cluboftheknobs.com