15V 7400 logic family??

[John Speth]

Dear Smartass - I mean ROmeo |:)

One of the important features I want in my sequencer is a presettable
sequence length control (last step control) that will do the following:
When in single mode, wait for a trigger pulse, step until the last step and
stay there (gate the clock off), and then wait for a trigger to restart it
again.  If a trigger happens mid-sequence, just restart at zero.  When in
continuous loop mode, free run and just automatically reset the counter
when the last step is reached.  As before, if a trigger happens
mid-sequence, just restart at zero.  I concluded that you'd need a counter
with a *synchronous* clear because you'd want the sequencer to wait at the
last step for exactly one clock cycle.  Whats more, I'm hesitant to use
anything else but a synch. counter because I want to minimize CD4051
addressing glitches that a ripple counter could possibly cause.  I think
the CD4022 is a ripple counter with asynchronous clear (from memory only -
no data book handy).

Somebody suggested using the CD4520, CD4516, or MC40163.  I understand the
MC40163 is a 15V CMOS version of the 74163.  I have an ancient CMOS data
book that doesn't have the CD4520 and CD4516 - I'll have to get a new one -
but these sound promising too.  One of the problems I've dealt with again
and again in my design is how to keep the sequencer start perfectly synced
to a trigger.  And I always end up needing a synchronous counter.  I guess
you can build a synch. counter from a ripple counter but you'd always be
one pulse behind because you'd need to step the ripple counter with the
clock and also clock in the ripple counter outputs to a D type latch to get
the ouputs stable.  Then you'd need to make sure the latch has a synch.
clear for starting at zero.  The 74163 does all this in one package.

(Does all this make sense or am I concealing some necessary details?)

At 11:58 AM 3/28/97 -0800, you wrote:
>
>Hate to be a smartass, but another way to do this is to use a CD4022 and a
>few other logic ics.  I'm looking at Roland's Model 182 sequencer from the
>System 100m line.   It's an 8 step design, but gives you the option of 16
>steps (serial mode) or 8 steps and two rows (parallel).  In parallel mode,
>you switch between rows to one output, in serial, just let each go to its
>own.  
>
>Anyway, hope I'm not making your project more confusing.  I just happen to
>be really into analog sequencers and have looked at a lot of different
>approaches.
>
>ROmeo


John Speth (johns at oei.com)
Object Engineering, Inc.
Vancouver, WA