Archive of the former Yahoo!Groups mailing list: MOTM

You can say that an ASR is sort of like an analog delay, and then quickly
add "not really," but in fact, that's ∗exactly∗ what it is. The difference
between the two is simply a matter of magnitude--the number of "stages" and
the speed at which they are clocked. An ASR is typically a very small, slow,
subset of an analog delay, and so it tends to be used for control voltages
rather than audio.

A single Sample & Hold module, as anyone who owns an MOTM-100 or -101 knows,
"samples" voltages varying at its input, and when a clocking pulse triggers
the module, whatever voltage was present that instant at it's input gets
reproduced at its output, and HELD at that level (even though the input
voltage may continue to vary) until the next clock pulse instructs the
module to hold a new sample.

An analog shift register is comprised of a few S&H stages in series,
typically three (at least that's what the Serge had, and some custom Moogs
I've read about). So what happens is that the same clock pulse is applied to
all three stages, and each time it is, the voltage is "handed down the
line." Stage 1 gives its voltage to stage 2 and grabs a new one, while
simultaneously, stage 2 hands off its sampled voltage to stage 3 and now
holds the "old" stage 1 voltage, and so on. All of this happens
[theoretically] instantaneously.

It helps to picture a line of three firemen, with a well on one side and a
fire at the other. Rather than each fireman running back and forth from the
well to the fire, they line up and "hand off" each bucket of water down the
line in a rhythm. This used to be called a "bucket brigade," and this name
is actually what was given to analog delay chips in the 70's! They used to
be called "CCD [charge-coupled device] 'bucket brigade'" chips!

The most famous (maybe because they were cheap and Radio Shack carried
them?) were the Reticon SAD-512 and SAD-1024 chips. Those numbers, believe
it or not, referred to the number of "sample & hold" stages in the chip! The
'512 had 512 S&H stages!

Since these were miniaturized, and marketed for audio delays, the capacitors
in these chips were of course tiny and didn't hold the charge very long at
all (they had high "droop rates")... they wouldn't hold a steady voltage for
a long time the way an MOTM-101 doeas. But that wasn't necessary because the
chips were being clocked at very high rates, in order that the fluctuating
voltages that were passed down the line could still be perceived as audio.

Anyway, I'm droning on, but I just wanted to point out to those that may not
know how analog delay units work that an analog shift register is, in
principle, exactly the same thing, but with less stages and optimized for
control voltages. By understanding the two together it might help understand
the ASR and start the gears turning about how you could put one to use. The
"canonic" use to control three VCOs may now become clear with the bucket
brigade analogy. (Picture each fireman singing a held note whose pitch is
equivalent to the amount of water in the bucket in his hand at that
instant... or is that just getting confusing again?).

Anyway, hope this helps.


Mr. T