Is it accurate to describe vintage Magnatone vibrato as pitch-shifting?

[Rhodesplyr]

I just finished a long, frustrating restoration of a non-working Magnatone M15A with a huge list of problems, so I became pretty familiar with the circuit. Its vibrato circuit was marketed as a true pitch-shifting circuit, but I wasn't sure that was what I was hearing, so I put it on a scope. I don't really see a pitch shift as much as I see shifting harmonic content from blending out of phase versions of the signal.

I work on Hammond Organs which have a pretty complex vibrato system (scanner vibrato), and I've looked at the output from that circuit on a scope. You can see that circuit modulating the length of a sine wave. Although the Magnatone circuit sounds very cool, I don't see that happening on the scope trace. It sounds just like all other Magnatone amps I've heard on youtube, so I don't think there's anything wrong the vibrato circuit.

Am I missing something?

[trobbins]

You are right that the Hammond and some other organs generate discrete notes that can have their specific frequency modulated by a low frequency - which effectively is like an oscillator that has its oscillating frequency varied, and equivalent to say a guitar string that has its resonating length varied by a finger.

To apply a frequency modulation to a signal passing through an amp, the common electronic techniques are that used by Magnatone, and by Wurlitzer. The modulation is signal frequency sensitive, in that the level of modulation is maximum with signal frequencies around 1kHz and drops off for lower and higher signal frequencies (1kHz is typical centre frequency for guitar use, but Wurlitzer used a lower centre frequency for its organ notes).

The ear hears the same modulation, but a spectrum analyser shows they are different - for example a 1kHz signal modulated by 10Hz will show as a 1kHz signal with discrete 990 and 1010 sideband signals from the Maggie, whereas the Hammond would likely show a flat topped peak from 990 to 1010Hz with steep sides (although I haven't done that spectrum measurement). So I agree that 'true' pitch shifting was a little marketing slight of hand.

[glebert]

It is pretty complicated to shift the fundamental frequency of a signal once it is already generated. Usually one would have to use something like a phase-lock-loop, at which point you are basically regenerating a signal and would lose a bunch of harmonic content from the original and it probably wouldn't work for guitar anyway with the wide range of inputs. Modulating a signal by mixing it with another tone is much easier. It is not that complicated to modulate an oscillator source to move the fundamental up and down, which is what I assume is what is happening with the Hammond approach.

[Helmholtz]

"Panning" between out of phase signal versions produces phase modulation. And phase modulation always means frequency modulation as well. The frequency shift is proportional to the rate of phase change. Those circuits are able to produce a frequency shift of 10Hz (Fender Vibrasonic) to 20Hz (Vox AC30), which might be hard to see on a scope using a 1kHz signal.

[Rhodesplyr]

It is not that complicated to modulate an oscillator source to move the fundamental up and down, which is what I assume is what is happening with the Hammond approach.

That's how it's done in transistor combo organs, but Hammond's mechanical tonewheel generator is locked to the AC mains frequency with a synchronous motor, so that approach is not possible.

Hammond's scanner vibrato system dates to 1949. The preamp drives a phase-delay line made up of a network of inductors, capacitors, and resistors. The first version has 25 inductors in series.

That phase delay line is then "scanned" by a rotary air capacitor with 16 static poles wired to the phase delay line and one rotating pickup geared to the tone generator. A switch assembly allows three different degrees of depth by scanning either part or all of the phase delay line, known to Hammond techs as the "line box."

A Chorus switch allows you to hear either the vibrato signal by itself or the vibrato signal mixed with the dry signal. The vibrato speed cannot be changed.

Edit: Hammond tweaked this circuit several times during the 1950s to simplify the phase delay line and lower parts count, eventually settling on the final version in 1957 that remained unchanged through 1975 and the end of tonewheel generator organs.

[nickb]

That's how it's done in transistor combo organs, but Hammond's mechanical tonewheel generator is locked to the AC mains frequency with a synchronous motor, so that approach is not possible.

Hammond's scanner vibrato system dates to 1949. The preamp drives a phase-delay line made up of a network of inductors, capacitors, and resistors. The first version has 25 inductors in series.

That phase delay line is then "scanned" by a rotary air capacitor with 16 static poles wired to the phase delay line and one rotating pickup geared to the tone generator. A switch assembly allows three different degrees of depth by scanning either part or all of the phase delay line, known to Hammond techs as the "line box."

A Chorus switch allows you to hear either the vibrato signal by itself or the vibrato signal mixed with the dry signal. The vibrato speed cannot be changed.

If the position of the tone wheel pickup were modulated by another variable speed motor with a crank to move it like a piston, that would also produce a vibrato effect. Then you'd have an all mechanical approach. Cool!

[nosaj]

If the position of the tone wheel pickup were modulated by another variable speed motor with a crank to move it like a piston, that would also produce a vibrato effect. Then you'd have an all mechanical approach. Cool!

Like this?

nosaj
http://modularsynthesis.com/modules/...er/scanner.htm

[trobbins]

Sort of like having a guitarist and combo on a moving train that approaches you (as the listener) and then stops and reverses back and then stops, repeatedly - aka a doppler effect in continuous cyclic motion.

The Maggie technique also introduces some tremolo modulation that is synced to the vibrato, which the other techniques do not include (afaik), which can set it apart. And as the level of vibrato depends on the centre frequency setting of the circuitry, effectively no vibrato effect is applied to low or high frequency signals.

[Mark Hammer]

Any chorus, flanger, or phase-shifter pedal can be transformed into a vibrato-type "pitch wobbler" by lifting/cancelling the dry signal that would normally be mixed in with the time-varied or phase-varied signal to produce the effect. A number of DIY hobby circuits, and commercial pedals, have implemented a non-varistor variation on the Magnatone circuit that yields the Magnatone sound.

In the current context, in which digital pedals will track entire chords and provide harmonies one or more octaves away in either, or even both, directions, it seems unwarranted boasting to characterize the miniscule deviations in pitch produced by the Magnatone circuit as "pitch shifting". That said, the vocal vibrato we view as an aesthetically pleasing and attractive embellishment of singing, is actually pretty minimal, and probably comes out of how singers try to maintain vocal pitch while sustaining a note.

As an aside, I made myself a stereo "Magnavibe", that I need to box up. Stereo Magnatone amps employ a single LFO that countersweeps the phase-shift subcircuits in each channel, by inverting the LFO output fed to one of those subcircuits. I used two independent LFOs, one for each phase-shift section, with global speed varied via a dual-ganged pot. The LFOs do not sync with each other, except for only momentarily and only now and then. I have to say that dual-independent is far more immersive than countersweeping. It's also less distracting and gimmicky since the pitch-wobble does not move from side to side. It just feels "busier", but not in a distracting way.

I guess the question to be asked is whether the sideband products noted with the spectrum analyzer are simultaneous, as they would be with a true ring modulator, or reflective of an average over time (i.e., one, then the other).

[trobbins]

A number of DIY hobby circuits, and commercial pedals, have implemented a non-varistor variation on the Magnatone circuit that yields the Magnatone sound.

Do you recall which circuits/pedals?
The varistors have been cloned by using semiconductor junctions, but they use the same circuit technique used by Magnatone (ie. a Moses style phase shifter).

[Mark Hammer]

Take your pick. Drawings and layouts of the Magnavibe are abundant. For the analog units, LDRs are used instead of varistors.

[R.G.]

As Mark points out, the Magnatione vibrato/tremolo effect is a variant of phase shifting. If you're a circuits kind of guy, you can look at the schematic and see that it's a phase shift stage much like a tube implementation of the Univibe shift stage: a phase splitter driving a resistor/capacitor phase stage mixed back into a buffer stage input. The tube implementation with varistors requires jumping through some hoops to get the 75-ish volts of change across the varistors to get them to vary their resistance, where the Univibe uses light change on an LDR. The electrons in the signal path don't know that the changing resistance is a varistor and not an LDR.
There are always side effects to getting a floating variable resistor. A number of the hoops jumped through in the Magnatone are the quirks of driving the varistors and the relatively huge modulating voltage on them that has to be wiped away from the signal by high passing and cancellation.
But it is just a phase shifter, with attached side effects.

[SoulFetish]

Do you recall which circuits/pedals?
The varistors have been cloned by using semiconductor junctions, but they use the same circuit technique used by Magnatone (ie. a Moses style phase shifter).

Anyone experiment using other SiC Varistors? I've seen some lower voltage variants sold on ebay (supposedly USSR stock).

[dmeek]

Here is the same idea using regular diodes. I experimented with it back in my analog synth building days.

https://music-electronics-forum.com/...1&d=1590429652

[trobbins]

I'd forgotten about the LDR's - Magnatone changed to using them in circa 1967. Ampeg used LDRs in two sequential vibrato phase-shift stages in the Gemini GV-22 guitar amplifier introduced in 1968. Forrest Cook similarly used an LDR vibrato back in circa 2008, and implemented digital LED waveform generation control which was pretty interesting.

I'm pretty sure the Magnavibe uses the Wurlitzer technique. Afaik, the Mindbender, TremO'Vibe, Tremster, Tube Wiggler, Vibromatic, Vibro-Stomp, VibraTrem, Vibrotron, and Vibravox also use the Wurlitzer technique.

Whereas I know the Surfyvibe uses the Magnatone technique but using diodes as per post #5's attached paper to achieve the required change in effective resistance but within a 9V battery useable swing, and jfets to replace valves, as we had some chit-chat while the designer was developing the circuitry.

The cat can also be skinned by using a variable reactance, rather than resistance, as per the Hammond L100 organ (AO-41 module) which used phase shift stages comprising saturable inductors, as well as mixing for chorusing, from 1961. Hammond continued using this through to the 1967 X66.

Afaik, Martin Manning has managed to get a useable vintage varistor characteristic out of just three modern metal oxide varistors of different voltage rating:
https://ampgarage.com/forum/viewtopic.php?t=23839