Re: freqmod step clock instability

[Ray <rehenry-at-up.net>]

On Thu, 09 Aug 2001, Patrick wrote:
> Ray
> Are there any specific variables you would like to include in the testing?
 
I think that the first question to answer is whether the jitter is a
constant with respect to period, max_velocity, and input and output scale. 
You might also want to consider cycle times for the motion modules.  

There is something else that you might want to consider as you begin to
get results.  Matt Shaver <mshaver-at-erols.com> talked with me during the
NAMES convention about his idea of granularity in output.  If I understood
him correctly there are output frequencies that are resonant or naturally
occurring with specific sets of the above variables.  These will produce
reasonably smooth pulse trains.  But there are also output frequencies
which are out of phase and cause the pulse train to bounce between slow and
fast sets of pulses. (Matt.  Help me here.)

Ray