Zappa feedback circuit explaination

[stompbox steve]

I have been wanting to post this for a while but when I saw a older topic talking about Franks circuit that was attached to a tone post, it was time.

Here is the explaination from  "Midget" Sloatman, (Zappa's chief guitar technician in the early 1980s). and a response from an engineer.  I lost tthe cointact with a change of job locations but thought someone where would be able to pick this up and run with it (as I am not able to)

Explaination from an old Guitar Player:
They were identical parametric filter circuits," explains"Midget .  One of the filters was set for the bass frequencies from about 5oHz to 2kHz, and the other one was set for the top end, from about 500Hz up to 20kHz." The filters had a variable resonant frequency ("q") independent from the EQ gain. "You could find a tone and get right on top of it, tweak it. and nail it," says Sloatman. The Q ranged from .7 to 10, or a very wide dynamic range to a very narrow one, and was adjustable via a 1/4" screwdriver notch on the face of the guitar. This allowed Frank to control his feedback characteristics in any hall. He could basically tune his guitar to the room, find out how the room responded to the amplifier, and dial it up so he could have maximum control of the feedback. That was the whole point behind the equalizers.

Engineer's response:
Actually, this is a pretty simple concept.  It actually sounds a little crude, but must work suitably for the purpose.

The Q of a filter refers to how sharply it filters around a given frequency (f-center).

A high Q filter (like 10), gives a VERY narrow signal response through it at
the f-center frequency.  That means it will pass only those frequencies that are
very close to the f-center of the filter.  It will knock down anything outside of this narrow band.

low Q filter (like .7) yields a very wide signal response, allowing lots of frequencies around f-center to pass through. 

The adjustment would allow this band to be widened or narrowed.

These are essentially bandpass filters, which can be built with a few low-noise p-amps and some resistors and caps. 

One bandpass is set to pass frequencies from 50 - 2000 Hz.  It would probably peak in the middle at around 1000 Hz.
When you adjust it, you'd be adjusting how much signal to pass for frequencies around 1000 Hz (above and below).

The other filter would do the same for 500 - 20,000 Hz, centered around 10kHz.

The Q adjustment would come from a variable resistor or cap (probably a cap in this case).

He was going to draw me a filter plot, but that's hard to do in an email tool but I never got back to them with work.  he suggested that even have an op-amp book with some example circuits.

Does anyone knwo where to pick up on this?

[Paul Marossy]

The first thing that comes to my mind is that one might be able to use a pair of Anderton Frequency Booster circuits or possibly a pair of Colorsound Inductorless Wah circuits to accomplish roughly the same effect. Both of these circuits could be built fairly compact so that they could fit into most guitar cavities.

[Mark Hammer]

While the Frequency Booster is on the right track (i.e., resonant but with high and low unaffected, unlike a wah), neither of them are high enough Q, actually.  The Frequency Booster is nice for goosing a distortion pedal to get more groan or grunt out of it, but can't produce what Frank used to call "ice pick through the head", and yields less selective boost than what you get diming a slider on the average graphic EQ.  I suspect nothing less than a state variable wil do.

Most parametric EQ circuits I've seen involve a state variable filter, though they tend to vary in terms of how much Q they can comfortably produce, how well that Q is preserved as one tunes up or down the spectrum, and how well it is preserved under cut conditions, compared to boost.  Of course, in this instance, we're interest only in boost so that aspect of a/symmetry of function is fairly moot.

[George Giblet]

> Does anyone knwo where to pick up on this?

There's nowhere else to go since it's "Actually, this is a pretty simple concept."

Two parametric equalizers.  That's it, it's all you need to know except perhaps choosing the amount of boost/cut (+/- 15dB is a good start).

[DDD]

Install magnetic sustainer into your guitar (for example STEALTH by Maniac Music, Inc) and you'll get fully controllable feedback regardless of the hall, power amps, your position on the stage etc, etc.

[Mark Hammer]

The goal is not feedback per se, but feedback at specific frequencies or notes.  Think of it like engineered wolf-tones.

[Paul Marossy]

While the Frequency Booster is on the right track (i.e., resonant but with high and low unaffected, unlike a wah), neither of them are high enough Q, actually.  The Frequency Booster is nice for goosing a distortion pedal to get more groan or grunt out of it, but can't produce what Frank used to call "ice pick through the head", and yields less selective boost than what you get diming a slider on the average graphic EQ.  I suspect nothing less than a state variable wil do.

Well, I did say it would have a similar effect, not be a replica. 

Alright, this topic comes up often enough here - who's gonna design one of these Zappa dual parametric EQs?!

[John Lyons]

I think paul is on the right track here. I have build the colorsound inductorless wah into a stompbox with only a frequency sweep knob. If you adjust the 33k to ground on the schematic you can get some very resonant tones. If you play around with the caps and resistors that make up the twin t filter here you can get close, or at least a nice resonant half %^&*ed wah /frequency boost. I built the two band parametric filter on the GEO site looking to boost frequencies to the same effect mentioned in the original question and I like the colorsound circuit better.
There may be a better way to do it but this one is sure simple.

John

[Mark Hammer]

aving built one years back, I can say with confidence that a parametric EQ built into a wah shell with the frequency pot footswpet makes a killer wah.  The nice thing is that unlike a wah, the treble and bass are always there.  What changes is the point of emphasis.

[Paul Marossy]

Hey, what if you tweaked the Tychobrahe Parapedal for a twin parametric EQ of sorts?! It uses a dual pot, so you could make them independent controls... 

[Paul Marossy]

I think I'm going to have to take one of my spare guitars and make it a testbed for this Zappa-ish EQ circuit thingie idea of mine... 

[Mark Hammer]

Well here's a circuit and board that should be quite small.  http://damien.douxchamps.net/elec/param_eq.php

[Paul Marossy]

So, I assume that the inputs of those twin circuits would be wired in parallel and then the outputs would be combined, right?

[Mark Hammer]

Not necessarily.  Parametrics are resonant sections, so you can cascade them without problem.

[Paul Marossy]

Parametrics are resonant sections, so you can cascade them without problem.

Really?! That seems counterintuitive, but you are more knowledgable than I, so I'll go with that. 

[Mark Hammer]

I forget where it's posted, but you should see if you can find the schematic for ules Ryckebusch's 4-band quasi-parametric that he designed for PAiA.  All cascaded.  If you look in the thread about 3-band EQ that was posted recently, you'll see a schem for a 3-band design from STM/Sebastian that is also 3 resonant sections in series.  So, nope, no real problem cascading them.

[Goatmeal]

Frank also mentioned a Green Ringer being built into a few guitars with the e.q. circuit.I think in his Les Paul and the Jimi strat,if I remember right.

[Paul Marossy]

no real problem cascading them

Well, you learn something new every day!