[sdiy] Fixed filter bank questions

mskala at ansuz.sooke.bc.ca mskala at ansuz.sooke.bc.ca
Sat May 11 14:52:38 CEST 2019


On Sat, 11 May 2019, ASSI wrote:
> was one that would ask for some value and tolerance specification, what E
> series you were having parts for and then spit out which combination of
> resistors would give you that value, either with the the least number of
> parts, or (if you'd put in a tight tolerance) the closest approach to what you

I have something similar online at
  https://files.northcoastsynthesis.com/spcalc/

It goes up to three parts, which is plenty to get within better than 0.1%
of a target even when they're E6 parts.  It also calculates based on the
nominal values, so if one wanted to hit a desired value *with a specific
tolerance* and not have to measure and select, it'd require going a bit
further.

I don't see how one can guarantee hitting a target value to better than
the tolerance of the parts used, without measuring them; if we imagine all
parts at the bottom of their range, or all parts at the top, it's easy to
show that in those two cases the entire circuit value is scaled up or down
the same percentage.  A scheme that doesn't include measuring the parts
ends up being forced to make assumptions about the distribution of the
parts' values (like the iid assumption you discuss), and then also hoping
not to randomly hit a bad case.  This is the same problem that was
instrumental in the mortgage-backed-security crisis:  combining high-risk
loans into low-risk securities depends on there being no, or no more than
an assumed amount of, correlation between them and when that assumption
failed, the calculated properties of the combination were no longer true.

-- 
Matthew Skala
mskala at ansuz.sooke.bc.ca                 People before tribes.
https://ansuz.sooke.bc.ca/



More information about the Synth-diy mailing list