[sdiy] ...dumb question (CD4017 sequencer)

[Robert Kent]

On Sun, 29 Sep 2002, Electronium wrote:
> I'm building a very simple 8 step/2 stage sequencer,
> using a CD4017 and here's what I can't quite figure out:
> How do you get the sequencer to stop so that it -
> A) does not put out the voltage of the current step and
> B) always starts again on step 1, when you restart it?

  I spent some time trying to design a gate sequencer with the
4017, and encountered these problems. Here's the closest I came
to a solution.

  I wanted to be able to run two chips in parallel (2x8 steps) or
cascade them for one long sequence (1x16 steps) and control the
length of the sequence either way (2xN, 1xN), all at the flick of
various switches. Cascading the two sequences introduced the problem
that when the first 4017 cascades to the second and is reset, it
triggers step 0 again (naturally). Since I was only using 8 outputs
of the 4017, I tied the 9th step (output 8) to INHIBIT, using it to
"park" the counter outside the visible 8 steps. Hitting RESET returned
the counter to 0 and the sequence started anew. This worked well for
an alternating sequence of 16 steps, but not so well for a shorter
sequence length since the counter would never reach the park spot.
So, I bumped the active outputs up to 1 through 8, leaving output 0
for parking the counter out of sight. Unfortunately, this meant that
when running the counters in parallel I always had an unavoidable
"invisible" extra step in the sequence.

  The longer I worked on this thing, the more complex it got: I soon
found myself adding inverters, buffers, pulse-length shorteners and
more 4017s for switching between counters. (Did I mention this was a
proof-of-concept design, hopefully intended to scale up to a massive
bank of sequenced switches which could be addressed as 8x32, 32x8,
and combinations in between?) Finding that the humble 4017 wasn't as
suitable for my grand design as I'd thought, I abandoned it entirely.
Right now some sort of digital control looks unavoidable (and would
allow a lot of crazy options like reversible and random sequences).

  Anyway, Ken Stone has a lucid sequencer design employing the 4017
that you might find helpful:

http://otherunicorn.0catch.com/synth/modules/gate_sequencer2.html

  Note that the 4017 comes up in an "insane" state when first powered
on; any sequence should start with a forced RESET rather than relying
on natural wraparound.

  Good luck!

--Robert Kent
  hanuman at ccsi.com