[sdiy] Where to buy ceramic caps

mskala at ansuz.sooke.bc.ca mskala at ansuz.sooke.bc.ca
Sun Dec 31 03:52:50 CET 2017


On Sat, 30 Dec 2017, David G Dixon wrote:
> And I have considered how sensitive the circuit is to precision.  That's
> what all the simulation is for.  It turns out that, as long as both the
> capacitors and the resistors  fall within 1% tolerance, it's almost as good

FWIW, I did some simulations a few months back of leapfrog-topology active
allpass filters for this application.  The leapfrog topology is *supposed*
to be optimally robust against component variations under certain
assumptions, but I found that it actually wasn't as good in this case as
just using biquad sections.  I think the reason is that first of all, the
robustness is related to interaction between poles at not too widely
separated frequencies, and the poles in a frequency shifter's Hilbert
transformer are not well laid out to take advantage of that; and second,
unless you go to great lengths to match resistors in the subtraction
circuits before the integrators, there's a loss of accuracy there which
introduces more error than is saved by the robustness of the leapfrog
topology.

The polyphase circuit sometimes used in ham radio for 90 degree phase
shifting (for sideband AM, etc.) is fun to look at, but:
  1. it's not especially robust to component variation (one early article
     made a bunch of false claims about its robustness which have been
     uncritically quoted a lot since);
  2. it requires twice as many RC circuits at whatever level of
     component precision you're using, compared to others; and
  3. it is far from having a flat amplitude response.  The hams don't
     usually care about that, but I think synth users would.
For those reasons I don't think it's a realistic choice for use in a synth
frequency shifter.
-- 
Matthew Skala
https://northcoastsynthesis.com/


More information about the Synth-diy mailing list