Unusual quadrature phase shifter

[Don Tillman]

For those of you into 90-degree phase shift networks for frequency
shifters and the like, check out the latest EDN magazine, June 5, the
Ideas For Design section, page 111. 

I've never seen a phase shifter circuit like this before.  Some sort
of quadruple helix thing.  It's weird as all hell.  

No, I'm not doing an asciimatic (now that's a challange!) but
basically it's two opamps providing a differential +/- drive to this
hairy cylindrical arrangement of 24 resistors and 24 capacitors, four
outputs from that go to two opamps wired as instrumentation amps with
differential inputs.  The opamp parts of the circuit are dirt simple,
there's no feedback around the Rs and Cs or anything, so it's
basically a passive RC circuit with input and output differential
buffering.

The author claims less than 1-degree of phase-shift error and less
than 0.2dB of magnitude error from 300 to 3500 Hz with 5% components.

  -- Don