Moog sequencer

[Grant Richter]

>I just visited some web page about the Moog modular, and was puzzled by
>a description of the sequencer:
>
>"It's possible for more than one stage to be on at once, actually. One
>way to make this 
>happen is to patch the output
>of one stage to the input of another. To get back to only one stage on,
>push a set button."
>

You can build a sequencer from a chain of mono-stables or a voltage multiplexer.
Most sequencers are built out of a counter and a decoder. Because of this only one
stage can be on at a time. If you build a sequencer out of a shift register, then if
you are only shifting one bit, it looks like a regular sequencer. But if you shift two
bits it does some thing new. The PAIA sequencer fro the 2700/4700 was built like
this from LM3900 logic stages (I think). There was a clear line that forced all stages
low and the the first stage would be "set" and then shifted down the line by the clock.

There was some logic in there so that whenever the first stage was set, the rest of
the stages would get cleared. That way you chained the gate output of the end stage
to the "set" input and that controled your sequence length. This logic could be disabled
so you could "set" the first stage while the bit still in there shifted out to the end
(two bits shifting - a two bit sequencer!). This would have the effect of adding automatic
transpose to a sequence.

It sounds like the Moogs use the shift-register design, I think the EML 400/401 does
also as I seem to recall a "set" input on that too. I think any sequencer that can have
all stages off must use that type of design? A counter with a decoder always has to be
decoding some state so they can not be cleared altogether.

Just my $0.0185 worth (inflation)