Green Ringer "Nulling" Mod

Started by Eb7+9, October 25, 2003, 05:12:56 AM

Previous topic - Next topic

R.G.

N.B. Re full wave rectifier octaves:
The closer you can get to an ideal full wave rectifier, the better the original signal is "nulled" out. A >half< wave rectifier creates significant second order products, but will never null the fundamental - if you consider the slew of harmonics and non-harmonics that come out of a guitar string and pickup as having a fundamental.

Most discrete rectifiers have two major defects, these being non-matched conduction characteristics and the offset turn-on voltage. The GR, Univox Superfuzz, Fender something-or-other all use slight forward biasing of the non-matched, non-ideal diodes or base-emitter junctions to try to make up for the offset.

Another way to do this which corrects both issues is to use an opamp based full wave rectifier. There's still tinkering to do, but it at least offers hope of getting to a closer-to-ideal full wave rectification.

But it adds a dual opamp, and is not as notionally simple as two diodes or two transistors to do the rectification.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

Eb7+9

#21
funny - last night I came up with what may yet be a much better balancing mod for the Green Ringer ... I now go the other way with driving point impedances

I also figured out how to make the jFET DOUBLER work much better ... easy fix, with a nice little schem if anybody's interested in trying

indeed, to some degree they are all just different kinds of FW rectifiers

Fourier's theorem assumes periodicity in its definition - ie., an isomorphism from cycle to cycle ... so, matching successive lobes in the FW transform simply leads to establishing a new isomorphic periodicity ...

Mark Hammer

I think, but am not absolutely sure, you're saying, or at least hinting at, the unpleasant reality that not only are the components of a discrete FWR not always what the label says they are (which I hope I conveyed earlier), but the challenge of producing an audible octave-up is made even more challenging by the uncomfortable reality that a guitar string is not an oscillator with steady pitch, waveform, and amplitude.  One millisecond of the signal is not going to look absolutely identical to the subsequent millisecond.

I tried to make the point earlier that the peaks of the rectified signal need to be of equal amplitude (or as close to it as feasible) for us to hear it as a doubling in frequency/pitch.  But you make a fair point that we can't necessarily rely on the input signal to provide us with identical successive waveform peaks.

Here's a thought experiment.  I know that when I wear out my strings against the too-tall frets on one of my guitars, I see "beats" in the string vibrations, resulting from the deformations of the string.  Such aberrations/deviations in the strings vibrating behaviour would be expected to result in even less resemblance between successive peaks in a waveform of a plucked string.  This would suggest that it becomes VERY difficult to get decent octaving off a worn and deformed string, even with the best of rectifiers and "matched" half-waves.

I wonder if that's true.

Eb7+9

#23
Mark,

I think what you're saying is right - we tend to roll off our highs to enhance the perception of octaving, but at the same time all that does is tame much of the upper intermodulations caused by string harmonics - which translates into a bunch of noise ... from a S/N point of view the octave sounds stronger - but this does nothing to shift the balance between carrier and second (produced) harmonic

octave enhancement otoh is really about getting rid of the fundamental, aka "carrier suppression" - and this can never be made perfect, there is always some residue ... the goal is to try to make it small ... the basis for that first Green Ringer mod

but that's only looking at things in terms of single-tone inputs ... there's more to an octaver than what it does to a single sinewave

give BOG a listen, and check out the part where Mayer's Octavio really starts doing "its thing" ... to me that's the more important side of non-linear circuits, it's there that we notice the effects of good matching or balancing ... for example, I discovered this long ago when matching parts (hFE) in the octaving section of the Shin'Ei "SuperFuzz"

---

as for designs,

the op-amp ideal-rectifier circuits will give you the most upper-fizz ... and are only good as reference or too-weird applications

http://www.lynx.net/~jc/superFullWave.html

using single-ended FW rectifier circuits, especially those implemented with jFETs and Triodes, naturally provide a way of obtaining a much cleaner "theoretical octave" ...  at least in the upper harmonics dept. ... even more so if there's a smoothing cap in the bottom part of the FW circuit







Steben

Quote from: Mark Hammer on October 15, 2020, 08:33:27 PM
I think, but am not absolutely sure, you're saying, or at least hinting at, the unpleasant reality that not only are the components of a discrete FWR not always what the label says they are (which I hope I conveyed earlier), but the challenge of producing an audible octave-up is made even more challenging by the uncomfortable reality that a guitar string is not an oscillator with steady pitch, waveform, and amplitude.  One millisecond of the signal is not going to look absolutely identical to the subsequent millisecond.

I tried to make the point earlier that the peaks of the rectified signal need to be of equal amplitude (or as close to it as feasible) for us to hear it as a doubling in frequency/pitch.  But you make a fair point that we can't necessarily rely on the input signal to provide us with identical successive waveform peaks.

Here's a thought experiment.  I know that when I wear out my strings against the too-tall frets on one of my guitars, I see "beats" in the string vibrations, resulting from the deformations of the string.  Such aberrations/deviations in the strings vibrating behaviour would be expected to result in even less resemblance between successive peaks in a waveform of a plucked string.  This would suggest that it becomes VERY difficult to get decent octaving off a worn and deformed string, even with the best of rectifiers and "matched" half-waves.

I wonder if that's true.

Frankly, pure octave up is out there. Digital effects are omnipresent. Look at the POG. What an effect. From octave to church organ sounds.... it delivers great performance and pure reproduction.
To me, a green ringer or other octavia(o)s stand for a stuborn effect with the usual ring tones. It offers the same stepping stones a fuzz face does to players used to gentle predictable overdrives. In a way, it is a sh"tty octave up, yet it delivers more as well.
  • SUPPORTER
Rules apply only for those who are not allowed to break them

Eb7+9

#25
Quote from: Steben on October 17, 2020, 09:47:01 AM

In a way, it is a sh"tty octave up, yet it delivers more as well.


octave up is a very idealized label ...

in the outside world, circuits like the Octavio, the Green Ringer and the Shin-Ei SuperFuzz, etc etc, are known as Full-Wave Rectifiers ... some have a twist thrown in to shift the harmonic profile - making them subjectively less sh"tty


Steben

Quote from: Eb7+9 on October 17, 2020, 04:26:41 PM
Quote from: Steben on October 17, 2020, 09:47:01 AM

In a way, it is a sh"tty octave up, yet it delivers more as well.


octave up is a very idealized label ...

in the outside world, circuits like the Octavio, the Green Ringer and the Shin-Ei SuperFuzz, etc etc, are known as Full-Wave Rectifiers ... some have a twist thrown in to shift the harmonic profile - making them subjectively less sh"tty

Yes, a bit like a little gating on fuzzes is not completely bad.
Fact is, I really like that ringy wacko tone coming from low gain settings. Those Band of Gypsies (album) sounds are sooooo sexy. But Jimi was a master in making theoretical circuit disadvantages work for him.
  • SUPPORTER
Rules apply only for those who are not allowed to break them

Eb7+9

#27
Quote from: Steben on October 18, 2020, 01:08:03 PM

Yes, a bit like a little gating on fuzzes is not completely bad.

Fact is, I really like that ringy wacko tone coming from low gain settings. Those Band of Gypsies (album) sounds are sooooo sexy. But Jimi was a master in making theoretical circuit disadvantages work for him.

careful comparing a poorly operating FW rectifier against a poorly operating fuzz to establish a theory ...

I was hoping somebody would notice I'm referring to an aspect of the interactions in terms of Jimi's note choices in some sections ... specifically, what happens when playing double-stops ... an acute minded guitarist might notice what happens to this key performance aspect when the FW lobes are well matched ... and with enough tinkering realize this applies to all of the FW circuits ...

it's something bench testing alone doesn't really afford, you need an axe in your hands to explore this with ...

Mark Hammer

Correct.

Indeed, the Green Ringer gets its name from the ring-modulator-like sounds produced by double-stops. Play one on any FWR-based octave-up and you;ll hear descending notes even though one is bending string up in pitch.  It's one of those quirks.  JC and I discussed this many years back.  He understands the math behind it better than I do.  ;) or likely ever will.   :icon_lol:

Eb7+9

#29
more an observation about the sensitivity of IM ringing to the matching of each side which I first noticed when cloning the Shin-Ei SuperFuzz ... with matched hFE's and all to bring out the octave ...

I'll be frank, over the last few hours I've been running simulation and found that my NULLING mod only applies with component mismatches occurring in one direction ... my oversight

Obviously, the best way is to start off by doing component matching in the FW section at the onset, especially the diodes ... sims show there's no need for a NULL control once proper component matching is established there ... the DPI imbalance that is present doesn't play a significant role either ...

viva-analog.com/danelectro-green-ringer-mods-redux-jcm2020/




Ben N

Quote from: Eb7+9 on October 15, 2020, 07:22:50 PM... so, matching successive lobes in the FW transform simply leads to establishing a new isomorphic periodicity ...
Simply?  ??? :icon_confused:
  • SUPPORTER

Steben

#31
Quote from: Eb7+9 on October 18, 2020, 02:45:13 PM

careful comparing a poorly operating FW rectifier against a poorly operating fuzz to establish a theory ...

I wouldn't call it a theory.

Yet one can't deny in the guitar world the list of "really commonly accepted useless" tones/pedals is very short. Vintage effects are already cherished by the fact they are vintage, let alone the quirks always attract.
Null modding a Green Ringer prooves this!  ;) It is perfecting a quirky pedal.
And we all love it.

Completey back on topic: that BIAS trimpot brings me back to the use of an opamp.
A BJT stage will always have assymetry, with full drive and fully used decoupling cap even more. Depending on values used, this could cast a shadow on cap and diode characteristics.
  • SUPPORTER
Rules apply only for those who are not allowed to break them

Mark Hammer

I will add that a back-to-back diode pair to ground on the output, with parallel cap, to foster a relatively constant volume does much to improve the audibility of the octave by limiting peaks, as well as trimming back some frizz.  That strategy is used by the Foxx Tone Machine, Fender Blender, and Superfuzz, although they situate the diodes after the octave-extraction and before the tone control and gain-recovery stage.  People mistakenly think those diodes are crucial to getting a fuzz, but one only has to lift them to find that there is still plenty of fuzz (at a much higher volume).  Dino/digi2t has also recommended inserting a small resistance (somewhere around a few hundred ohms) between the diodes and ground to soften any clipping the diodes do (on top of the peak limiting).

Between these mods and what JC shows, we now have a "complete" and optimally-functioning Green Ringer.

Steben

  • SUPPORTER
Rules apply only for those who are not allowed to break them

Eb7+9

#34
Quote from: Steben on October 19, 2020, 06:45:05 AM

I wouldn't call it a theory ...


my point was about the gating you used to compare the two ...

as I'm hoping you'll notice, it's the relative gating that we can play with here - by upping the front-end gain, and re-dropping at the output ... doing a dynamic pre-emphasis of sorts ...

of course, there's one other trick we can play with there ;)

like I said, the NULL mod works if you've got significant mismatching going on, and you happen to have that mismatching cause an offset going in the right direction for the MOD to work ... obviously, it's better to get rid of these offsets at the onset when building and not have to rely on the NULL mod later ...

Steben

Would a reverse connected stereo (dual gang) control reduce some of the unidirectional problems or will this affect the overall performance?



  • SUPPORTER
Rules apply only for those who are not allowed to break them

Mark Hammer

There's a lot that can be mismatched, though.  Not just at the phase-splitter, but after as well.  For instance, do the diodes have the same forward voltage?  Are the 68k resistors to the bias point equal in value?  I suppose one might have some concerns about the similarity of the two caps in series with their respective diodes, but I would put that way down the list, given that their hypothetical difference would really only have impact at the lowest end of the pass-band.

I've noted in past that, in his discussion of single-transistor phase-splitters, Bernie Hutchins notes that the two opposite-phase copies of the signal coming off the emitter and collector, using equal-value emitter and collector resistors, may not be identical in amplitude.  I suppose, in a perfect world, one uses a multi-turn trimmer for the collector resistance, and tweaks it until you see equal-amplitude copies of the signal on a scope.

But this is about rock and roll, so the question is really what sort of simple mods get us to the sound we want with the least effort and requisite equipment.

Steben

Quote from: Mark Hammer on October 20, 2020, 08:07:52 AM
There's a lot that can be mismatched, though.  Not just at the phase-splitter, but after as well.  For instance, do the diodes have the same forward voltage?  Are the 68k resistors to the bias point equal in value?  I suppose one might have some concerns about the similarity of the two caps in series with their respective diodes, but I would put that way down the list, given that their hypothetical difference would really only have impact at the lowest end of the pass-band.

I've noted in past that, in his discussion of single-transistor phase-splitters, Bernie Hutchins notes that the two opposite-phase copies of the signal coming off the emitter and collector, using equal-value emitter and collector resistors, may not be identical in amplitude.  I suppose, in a perfect world, one uses a multi-turn trimmer for the collector resistance, and tweaks it until you see equal-amplitude copies of the signal on a scope.

But this is about rock and roll, so the question is really what sort of simple mods get us to the sound we want with the least effort and requisite equipment.

Emitter follower is always lower than 100 %, collector is 100%..... in theory
  • SUPPORTER
Rules apply only for those who are not allowed to break them

PRR

> 100 %, collector is 100%..... in theory

It's not that simple.

It occurs to me that while the cathodyne is much better balanced than you would think, for "clean" uses, balance may be bad when driven or loaded hard. The apparent different impedances E and C actually balance in the linear zone but not in overdrive.

I could point out that opamps are 19 cents a pair.
  • SUPPORTER

Steben

Quote from: PRR on October 20, 2020, 12:59:17 PM
> 100 %, collector is 100%..... in theory

It's not that simple.

It occurs to me that while the cathodyne is much better balanced than you would think, for "clean" uses, balance may be bad when driven or loaded hard. The apparent different impedances E and C actually balance in the linear zone but not in overdrive.

I could point out that opamps are 19 cents a pair.

Although being 200% correct and though I advocate using an opamp as input stage, as an "early on" pedal, signals do not reach above a serious drive region at the phase splitter.
If we add Mark's suggestion of clipping (I suggest soft) before the splitting stuff, the cathodyne is far from being driven.
  • SUPPORTER
Rules apply only for those who are not allowed to break them