MOTM

There are 2 ways for a reset function to operate - synchronously or
asynchronously. Those are just fancy schmantzy terms for saying it either
jumps to step 1 as soon as the reset signal is applied (asynchronous
operation), OR it jumps to step 1 on the next clock pulse AFTER the reset
signal is applied (synchronous operation).

Synchronous behavior is more intuitive to the user for programming
sequences - take the simple case of wanting to reset to step 1 after any
arbitrary step x. You just patch the step x stage output to the reset input.
It will signal a reset as soon as step x becomes active, but the switch to
step 1 will occur on the next cycle, which is what you want. For more
complex cases where several signals must be set in order to branch
somewhere, just set them up on the previous step, and your branch occurs on
the next clock, exactly in rhythm.

This is well known standard synchronous logic counter behavior. Using a
microprocessor, Paul can emulate synch OR async behavior in software. Or
even user switchable, but that level of flexibility is probably too
confusing.

I can't imagine that the SOS review is accurate on this point. If most
sequencer designers were committing such basic logic errors such that step
one is not played first, they would be out of a job PDQ!

Dave Bradley
Principal Software Engineer
Engineering Animation, Inc.
daveb@...

> >
> > I haven't any experience of analogue sequencers yet, but they describe
> > in the review the fact that most sequencers RESET function resets the
> > sequencer to step 1, which means that the first note played will be step
> > 2. The concussor resets to step 16, so step 1 will be the first note
> > played. Seems like common sense to me, but they say it is the only
> > sequencer that does it.
> >

I fully agree with both of Dave's points below. I can imagine a homebrew 
sequencer that resets to stage 1 so that the next clock pulse (and gate 
signal for the EG) would advance it to stage 2, but I can't imagine 
manufacturing (or buying!) a sequencer like that: this would merely be an 
example of poor (i.e., un-MOTM) design.

As far as an asynchronous reset (or any other asynchronous logic circuits), 
I'm not sure I would like that at all due to the way I use sequencers. Has 
anyone found this type of logic useful for EM?

But Andy's original post made me think that it might be useful to have a few 
RESET inputs, each for resetting to different stages. After which I thought 
it might be better to just have several PULSE inputs, each of which could be 
user defined for RESET TO X, LOOP TO X FOR N TIMES, BRANCH TO X, HOLD UNTIL 
J, etc.

And I'll leave the trivial details of programming this to Paul.

JB

In a message dated 8/22/99 12:27:24 PM, daveb@... writes:

>This is well known standard synchronous logic counter behavior. Using a
>microprocessor, Paul can emulate synch OR async behavior in software. Or
>even user switchable, but that level of flexibility is probably too
>confusing.

>I can't imagine that the SOS review is accurate on this point. If most
>sequencer designers were committing such basic logic errors such that step
>one is not played first, they would be out of a job PDQ!