From gstopp@fibermux.com Mon Jan 20 15:04:29 1997
    20 Jan 97 18:03:51 +0500
    20 Jan 97 14:48:26 +0500
          id TAA49832; Mon, 20 Jan 1997 19:34:46 GMT
          id TAA88993; Mon, 20 Jan 1997 19:34:42 GMT
Date: Mon, 20 Jan 97 11:38:50 PST
From: gstopp@fibermux.com
Encoding: 57 Text
Subject: Re: Keyboard Interface - help needed

     I've built the EN#45 interface, long ago....
     
     Actually the design is not flawed, just kind of non-robust. It's very 
     prone to contact bounce triggering errors. The one I built (in 1975!) 
     worked but had the nasty habit of triggering the S/H upon key release, 
     which of course meant that the held voltage was the wrong one.
     
     In frustration I then built the digital design from EN#68, which works 
     flawlessly under all playing styles. I left out the trigger generation 
     part, however, since it has a pretty high chip-count, and besides I'm 
     used to having just a gate with no trigger anyway.
     
     Over the years I've used the same basic circuit several times. I 
     usually copy the circuit as far as the counter/de-mux/latch logic 
     goes, and then add my own re-triggerable monostable to produce the 
     gate. One time I did want a trigger circuit (for a selectable 
     re-trigger/non-retrigger design in a five-octave Minimoog) and for 
     that I used an op-amp high-gain differentiator on the CV to produce 
     the multiple triggers.
     
     Probably the circuit in EN#68 can be updated and simplified a little, 
     these days. However IMHO there are just too many benefits to using a 
     digital interface to produce a monophonic analog voltage from a 
     physical keyboard. The keyboard function is basically a digital one in 
     the first place - a bunch of switches to be closed and opened (a 
     binary thing), with one and only one closure taking priority (a logic 
     thing), with the last closure latched (a latch thing), converted to an 
     analog voltage (DACs are easy), plus a single bit that indicates 
     whether any switches are closed (the gate output, a boolean OR thing). 
     The latched DAC is way better than any analog sample/hold, because it 
     has infinite hold time with no drift or droop.
     
     So my advice is to dive into the digital approach -it's worth the 
     trouble.
     
     - Gene
     gstopp@fibermux.com


Subject: Keyboard Interface - help needed
Author:  POLARIS@VAX1.Mankato.MSUS.EDU at ccrelayout
Date:    1/19/97 3:52 PM
     
I'm trying to make a keyboard interface from Electronotes #45, which is 
an analog design, and am failing miserably.  The CV gets generated just
fine, but does not get to the CV out due to lack of triggering.  The basic 
design of the circuit seems to be flawed, but I assume it worked when it 
was published.  I would like to talk in-depth about this with anyone "in 
the know", or about alternative ways of generating gate and trigger signals 
from single-contact Pratt-Read keyboards.  I am hesitating to go with the 
digital scanning scheme, but I may be forced to if the analog way fails to 
produce acceptable results.
     
I can't get any further on my modular project until I get this out of the 
way, so your help will be highly valued.