AW: Harald Bode Barberpole Phaser Questions

Haible Juergen Juergen.Haible at
Fri Jan 16 09:41:40 CET 1998

	>1) Anybody have any ideas for the construction of the all-pass
filter that
	>is used in series (or in parallel) with the dome filters in the
	>shifter? The allpass filter is used to add notches to the phase
shift effect
	<in the audible areas of the spectrum, around 300 to 1500 Hz.
Any idea what
	<resistor and capacitor values should be used in the circuit?
Should the
	<values be the same for each op-amp stage, or should they be
staggered in

Any phaser would do. Optimum would be if you could use
the all pass filter output separately (without the direct signal
mixed in), but the mixed signal is interesting too, and the effect is
only slightly different. 
Haven't tried staggered poles yet, but it's easy to try it on the
weekend, as I have a variable slope filter/phaser, so I can
simulate staggered capacitor values with one control voltage.
But as the effect is there even without an external phaser, I am
sure everything in between will work as well.

	>2) Is it necessary to have a fully functioning, lab-quality
	>shifter for this effect, or is it possible to cut corners if
all you want is
	>a barberpole phasing effect?  Do you need such a complicated
setup for the
	>dome filter, or could you construct a dome filter with less
stages and less
	<precise components? How about substituting a manually
adjustable >quadrature
	>function generator for the quadrature through-zero VCO in the
Bode patent?
	<And how about using cheaper OTA's for the multipliers?  If you
keep the
	>quadrature oscillator frequency below the audio range, would a
	<circuit be needed?  Is the precise nature of the circuit
essential to having
	<the small shift amounts needed?

Very good idea. I love building high end stuff, and then extracting
low budget circuits for special applications!
That's what I think (without having tried it):
I'd rather keep the dome filter wide band (12 pole) in order
to avoid too large amplitude modulation effects at extreme
frequencies. But probably you don't need to choose 0.1%
components anymore. Should even work with 5%, I guess.
I'd omit the thru zero capabillity of the QVCO - otherwise my
circuit should be very cheap. (No more FWR with low offset opamps)
The modulators: You can hardly go cheaper than I did with
MC1496 chips. For this special application, you might be
able to omit most of the trimmers, do the waveshaping with
the "cheap" upper input (would still give < 3% distortion on the
sine shape), and use the lower, linearized inputs for audio.
You won't need a compander system anymore: You don't hear
the carrier, don't care for remains of side bands, and the lower
input of the 1496 is not noisy. Hey, I *love* this idea !

	> P.P.S.  I love the term "Klangumwandler" for Frequency

It doesn't sound _that_ elegant to German ears, believe me.
"Sound converter" - this could be everything.


More information about the Synth-diy mailing list