Csound Csound-dev Csound-tekno Search About

Re: Phase shift by +-90 degrees?

Date1998-11-28 08:09
FromSean Costello
SubjectRe: Phase shift by +-90 degrees?
Job van Zuijlen wrote:
> 
> I would try to get some information on how this is done in the analog
> world.  I used to know this 20 years ago :-|
> 
> There are applications for which you need a sine as well as a cosine
> output over a fairly wide frequency range.  This easily done digitally,
> but for an analog application you would want to use one sine wave
> oscilator and than phase-shift the output 90 degrees. Older analog
> implementations of surround sound implementations must have had similar
> circuitry.  I would search for patents that describe those.

Analog frequency shifters (single sideband amplitude modulation) relied
upon having an input signal split into sine and cosine components over
the audio frequency range.  Frequency shifters like those made by Harald
Bode used a filtering scheme called a Dome filter, which relied upon two
parallel allpass chains that would output signals that were 90 degrees
+- a small amount of error from 15 Hz - 15kHz. The circuit required
precision resistors and capacitors.

A similar analog circuit is detailed at
http://www.synthfool.com/diy/hj_fs.html. This circuit is based upon the
Electronotes frequency shifter.  The site has the RC values needed for
each stage of the phase shifting network.

As far as I know, the most common way of implementing 90 degrees phase
difference networks in the digital realm is using an FIR filter as a
Hilbert transformer. However, if anyone gets a Hilbert transformer
working using the allpass network topology, please let me know.  I've
been trying to get one of these working (hence my questions on phasers),
but I haven't been able to figure out how to successfully construct my
own allpass networks from scratch in Csound.  Any help would be greatly
appreciated.  

Thanks,