MOTM

>>>>
Just how nuts would it be to DIY and build this
thing Dave? PCB's? Not that you would ever even be
willing, or offering, just curious ;)
<<<<

Reasonably nutty. It's dense enough that one or more pcbs would be 
necessary. I suppose one could wire-wrap a one-off, but I have only 
done a few wrap projects, and I never gained a taste for it. Wire 
wrapping would offend the MOTM build aesthetic, that's for sure.

Density of build is one reason I never attempted it. You would need a 
large sub-board whose sole purpose was to connect to all the switches 
and pots, and multiplex them. Then a daughterboard for the clock and 
shift register stuff.

Another reason is that for all its complexity, it is a fixed route 
design, like a Minimoog. You get 16x2 or 8x4 topography, period. You 
get clocked or CV control of stage selection, period. To expand it by 
chaining to another identical unit requires summing the outputs in an 
external mixer.

It is interesting how I came to switch my thinking to a modular 
sequencing approach. I wanted to add one more feature - the ability 
of the Serge TKB to clock the vertical position across rows as well 
as horizontal position. The design I had simply couldn't accept 
another feature - I had filled up a full rack width, and wasn't 
willing to sacrifice anything else for this feature.
 
The more I looked at the E-MU modular design - separate clocking, 
address generation, and cv generation modules - the more logical it 
seemed. Yeah, it's a space hog compared to an all-in-one box, but a 
single CV unit now has the room to utilize my nifty design of 
simultaneous 8x4, 16x2, and 32x1 outputs. I do understand the allure 
of the all-in-one box, however. It would be ungodly expensive just to 
use good panel hardware, though!

17 pots @ $5.00
17 knobs @ 2.00
31 jax @ $1.20
22 switches @ $5.00
18 leds @ 1.00
1 panel @ 50.00
-----------------
Total $453.20 yikes!

Not to mention pcbs, parts, labor, etc. You would have to charge 
$800 - $900 minimum to build one for somebody.

Moe