From obxwindsurf@yahoo.com Tue Oct 07 06:26:05 2003
Subject:Re: Next Generation Vibrato/Chorus Clone - reflections - I don't think so.

You are correct in that DC Bias disappears in the delay line as the
coupling capacitor from the output of the percussion amp and feeding
the delay line's input brings things back to a ground level. Also
the scanner is effectively an air capacitor so any selected tap of
the scanner is capacitively coupled to the preamp just before the
expression pedal.

I've only seen one reference of all the ones I've read, to the delay
line being discussed as a transmission line. You will only get
reflections if one or more of the inductors form an LC tank circuit
and resonance will occur. No such LC tanks exist in the Hammond
Vibrato line box, even with the 22k resistor (normally bypassed with
the vibrato switch), floating the inductor commons above ground. You
are also correct in saying "it probably behaves differently when
there is a resistor to ground". The following discussion explains
what that behavior is, according to well established electronic
engineering principles.

Each leg of the delay line is a second order low-pass filter. Delay
is probably a misnomer here as the delays are actually produced by
phase shift which is a sort of delay (phase lag) in the signal. With
a given pure frequency present at any given point in time the phase
lag between input and output can be interpreted as delay, but it
changes with frequency as phase shift through an LC circuit is a
function of frequency. Phase shift is cascadable, that is once the
phase shifted signal from output of one stage is present at the input
of the next stage, additional phase shift occurs. Since phase shift
is a function of frequency, that means that even with one drawbar
selected (a pure tone for the purposes of discussion) each key
pressed will exhibit a different degree of phase shift. If these
phase shifts are scanned very quickly such as is the case with the
vibrato scanner, the effect when heard is a "pitch wavering". When
phase shift is summed with the original signal, cancellations occur
when the phase of a delayed signal is equal and opposite to the input
signal. The effect is the chorus that everyone is familiar with.

Any resistor-divider networks present in the first 6 or so branches
of the Vibrato line are designed to "pre-attenuate" the signal.
Passive inductor circuits have losses. The thought of pre-
attenuating the signal is to compensate for losses and attenuation
that occur later in the vibrato line so that the attenuation of
signal varies little from one end of the line to another...but it
does vary. The whole thing can be modeled as a single impedance
(formed from the multiple impedances of the early stage voltage
dividers in parallel with the impedances of each of the LC circuits)
that is either ground referenced, or floated 22K ohms above ground
when the chorus switch opens across this resistor. This allows a
portion of the dry signal to be present in the Vibrato line as it no
longer has a solid reference to ground.

So in effect there is an >>impedance<< bias imposed onto the delay
line (that cannot be cancelled by capacitive coupling) when the
switch across R44 is open, that when fed into one half of the dual
triode at V4A causes the input signal from the output of V1 through
the voltage divider of R9 and R10 to mix together resulting in the
cancellations and the chorus effect to appear. This mixing takes
place at the junction of C8, C15, R27, R28.

If you look at the schematic, it indicates that R10 is selected and
used to balance the [dry and delayed] channels [which sets the depth
of the chorus effect].

If you were to do this with op-amps, you could still use the same
circuit, although you'd lose the "R44" resistor and instead switch in
the dry signal into one input of the op-amp mixer in the "chorus"
setting and leave that input unconnected in the "vibrato" setting.
Phase shift can be created in op-amp circuits either using 2nd order
low pass filters, or a voltage- or digital-controlled phase shifter.
The gains of the two inputs to the equivalent op-amp circuit (set by
the ratio of the feedback resistor to each of the two input
resistors) are adjusted to achieve the same effect (chorus depth).
Same effect, different technique driven by the differences in the
technology.

I'm sure if you engage an abstract mathematician the whole analysis
could come up with reflections and a transmission line model due to
inductive components in the capacitors and stray capacitance in the
wiring, but as is well documented, the design intent of the circuit
is to create multiple discrete phase shifts (they didn't have
continuously-variable voltage control to achieve this), and then scan
them up and down rapidly (via the scanner). The phase shift when
scanned back and forth creates a pitch variance perceived as vibrato.
Summing this with the original signal and the resultant cancellations
produces what we know as chorus.

It doesn't appear that any other motivation was behind it. The
circuit is relatively simple, all things considering.

Regards,
Kevin

--- In CloneWheel@yahoogroups.com, "elggobo" wrote:
>
> Regards
>
> James