Magnavibe question

[Kipper4]

Ok here's a schematic for a starter on Fred's page.

http://revolutiondeux.blogspot.pt/2012/01/bigfoot-fx-magnavibe.html

Now comes the dumb question......
Is the transistor set up with a positive or negative feedback loop (the one with the ldr and cap)
Is it possible to use an op amp and make it happen? Or is it a quirk of the transistor circuit that it can only happen with a bjt ckt?

Thanks
Rich

[Kevin Mitchell]

Check out the easyvibe
Schematic here

Hope that steers you in the right direction.

[Kipper4]

Thanks Kevin.
Both nfb and pfb then in the easy vibe.
Maybe I'll bread it up when I get back with a phase shift oscillator but I suspect it won't have that magnavibe feel. Sometime things are best left alone.
We will see.

[Groovenut]

Rich,

You essentially have a Darlington Concertina splitter with a low output impedance at the emitter and a high output impedance at the collector. The emiiter and the collector will have the same signal output but will be 180 degrees out of phase with one another. The LDR is therefore acting as a mix resistor for to out of phase signals. When the signals approach unity amplitude they cancel each other out giving very low output.

IMO, allegedly.....

Law

[Kipper4]

Wow Law.
I had to wiki that one.

https://en.m.wikipedia.org/wiki/Phase_splitter

Concertina spitter I feel a name coming on.
Fantastic stuff guys.

[nocentelli]

It looks like a single stage of an easyvibe-type phaser, but with two out-of-phase signals (one wobbled by the LDR) coming from collector and emitter, instead of sending the same signal to the out-of-phase inverting and non-inverting inputs (with one being wobbled by an LDR), so could presumably be replicated with a single opamp stage. It might be fun to try a dual opamp and have the other do the LFO, but I wonder if there might be an increased risk of ticking if the same chip is doing both audio and oscillator.

[samhay]

I did a stereo one of these recently using op-amps rather than BJTs: http://www.diystompboxes.com/smfforum/index.php?topic=111802.5

It can be difficult to get the vactrol to give a musical sweep and/or get enough of an 'effect' out of a single stage, but otherwise works well enough - no worse than a BJT stage anyway.

[Kipper4]

Thanks Leo and Sam.
I'll have to reread that a few times Leo to get my head around it.
I remember the ACE Sam. Did you ever do the pic chip version?
The magnavibe is so effective I'm not sure it's worth putting a lot of effort or parts into modernising it as such, but the theory of how it works and the potential for new effects using the technology excites me.
So I wanna dig a bit.
Some might say get a life. I say I already have a nerd life and I'm loving it.
That's all.

Feel free to add to my excitement.

Thanks guys

[samhay]

>I remember the ACE Sam. Did you ever do the pic chip version?

It worked on the breadboard beautifully. Then I built it, and it isn't quite right (works, but sweep isn't great), so I threw it in the pile of things to look at one day, and there it has lay.

[Gargaman]

Some might say get a life. I say I already have a nerd life and I'm loving it.
That's all.
Feel free to add to my excitement.

Man, I still don't dig the whole thing even with a fuzz or booster, but I'm very animated to see what you will discover.
It's an amazing effect, look simple and great.

[Kipper4]

What do you guys think of the possibility of making a dual magnavibe for that sea sick kinda thing Sam has going on with his ACE.
But using twin bjt concertina circuits like in the magnavibe (post1)
So can't I just make a std concertina version as post one and a second concertina but with the 100nf cap and ldr reversed so the two are out of phase. Use a single PSO to drive the vactrol led(s) and use a dual 25k speed pot and separate depth pots for both signals
Or would the two out of phase signals simple cancel each other out ?

Boom or bust.
Boom. I get a stereo magnavibe
Bust. I get a nada sound.

[blackieNYC]

Uh, wait a minute.  That Rev Deux magnavibe demo came from that schematic? It has 1/4 of the parts of an Easyvibe! What's the catch?  And that oscillator - it's sinusoidal, or at least smooth? 
That demo sounds awesome.

I have a straight tremolo, and a 4or6 stage phase royale. But I want to get into this Vibe vibe.

[nocentelli]

Uh, wait a minute.  That Rev Deux magnavibe demo came from that schematic? It has 1/4 of the parts of an Easyvibe! What's the catch?  And that oscillator - it's sinusoidal, or at least smooth? 
That demo sounds awesome.

The magnavibe is a single wobbly stage and no option for mixing in the dry signal, so you only get a relatively subtle vibrato effect and no chorus option. The LFO transistor has a slightly limited speed range (i.e. won't start or peters out at the faster or slower ends of the range). To get the pleasant shimmery vibrato sound, a medium/fast rate and modest depth works best, so the magnavibe does a couple of sounds quite well but is not terribly versatile. The easyvibe has four wobbled stages and a fixed input/buffer stage so a more complex swirl, and a chorus sound is available, and the LFO is stable over a wider speed range. You chain several magnavibe stages together for more depth, and use a different (opamp) LFO, at which point the parts count will shoot up meaning you may as well build the easyvibe.

[PRR]

> transistor set up with a positive or negative feedback loop

The input bias is Collector-Base negative feedback.

The interesting part, the C-Rvar network, is no kind of feedback.

(You can't have effective feedback from collector to emitter of the same transistor without some other leverage.)

The cathodyne/concertina is slick because it has two equal/opposite outputs to drive the C-R network. With opamps you want two; or use the two-input form Kevin posted.

> When the signals approach unity amplitude they cancel each other out giving very low output.

No. The voltages are always equal. At low freq the C won't pass so we get output from the R at input polarity. At high freq the C passes better than the R so we get the other polarity. In the middle we get essentially the same voltage but at a phase "between" zero and 180 degree. As drawn it runs from zero in bass to 180 at far treble. For R=1.6K the phase passes through 90 degrees at about 1KHz. Lower R makes the 90deg point higher freq, and vice versa.

A single stage of this is very subtle.

> Concertina splitter with a low output impedance at the emitter and a high output impedance at the collector

The collector is 7.5K which is "fairly low" compared to what it drives. There's no particular significance to this value.

> that oscillator - it's sinusoidal, or at least smooth?

It is the classic 3-R-C Phase Shift oscillator. Just like a Fender (but transistorized).

When adjusted exactly (no provision here), the output can be Sine with under 1% THD. At that adjustment it would take "forever" to start-up. Especially when tuned below the audio band. We always set it for gain 20%-100% higher than "exact" so it will start-up today (under a second). When it comes up it distorts itself to the "right" gain and self-limits at full output. The slick thing about this form is the high-pass from collector to base. This NFB tends to give clean wave at the collector, putting the distortion at the base (we don't use that signal). The base is going in and out of cut-off, but any sudden change at collector is C-C-C-ed back to base to smooth the collector waveform.

It could drive more than one LED, but at three red LEDs you may be running out of voltage under a 9V supply.

The obvious problem is three resistors must be changed for any large change of frequency. Changing just one resistor gives a limited range, and (as leo said) tends to go out of oscillation if you take it too far.

A problem common to all such things is that for "normal knob rotation" over more than a part-Octave of frequency, the speed pot must be Reverse Audio taper.

We often just trim one resistor with a Linear pot and accept the action being cramped at one end.

If you come across a 3-gang reverse-audio pot (ha!) of useful value, this type LFO sweeps nice over a 30+:1 range of freq. If you don't mind the knob turning "backward", 4-gang Audio pots used to be a thing.

[Kipper4]

Uh, wait a minute.  That Rev Deux magnavibe demo came from that schematic? It has 1/4 of the parts of an Easyvibe! What's the catch?  And that oscillator - it's sinusoidal, or at least smooth? 
That demo sounds awesome.

The magnavibe is a single wobbly stage and no option for mixing in the dry signal, so you only get a relatively subtle vibrato effect and no chorus option. The LFO transistor has a slightly limited speed range (i.e. won't start or peters out at the faster or slower ends of the range). To get the pleasant shimmery vibrato sound, a medium/fast rate and modest depth works best, so the magnavibe does a couple of sounds quite well but is not terribly versatile. The easyvibe has four wobbled stages and a fixed input/buffer stage so a more complex swirl, and a chorus sound is available, and the LFO is stable over a wider speed range. You chain several magnavibe stages together for more depth, and use a different (opamp) LFO, at which point the parts count will shoot up meaning you may as well build the easyvibe.

Ok.
Thanks First to Paul for the in depth explanation and your commitment to passing on your knowledge. It is appreciated. I'm in awe of your inate ability to look at a circuit its values and being able to do quick and dirty calc to work out what it's capable of and its limitations.
This is something I've only come across in the best of engineers be they electrical or mechanical.
You are a lucky and good man. So thanks.

Right Leo.
I Could try cascaded magnavibe said for more depth. Maybe when I breadboard it I will.
Since the lfo only only behaves when at med to fast rate.
Would it be possible to put a series resistance between the speed pot and ground or would it upset the lfo too much and it starts to play up?
I'm sure during the course of my investigations I will bread the easyvibe. Try swapping out the lfo, maybe an op amp type lfo, something more stable.
I wonder how two parallel magnavibes in one box would sound, left and right output to differant amps and both set to differant settings, rate, depth. Could sound horrible. Time will tell.
I don't know what it is about this circuit. I just liked to samples, videos I've seen of it. That combination of tremolo and vibe is just great.
Thanks Leo your continued support is important and appreciated.

I'm gonna go look for that vid I was watching on you tube last night.

[Kipper4]

On second reflection this is not what I thought it was and could probably be done with the 4ms panner. (Twin temelus lunes) I'm not hearing the pitch modulation I thought I was last night.
Feeling tricked.


http://youtu.be/ouw7xuRe15A

Think again.


And IIRC there's already a series resistor to gnd on the rate/speed pot so ignore that too. Although I read a mod to change that on Madbeans forum.

[Kipper4]

It's that wow and flutter that I really like. Maybe. It's time to get my old akai reel to reel out of the loft and dust it off.

[samhay]

The magnavibe can do that - as other's have said, it is a bit of a one trick poney, but if set up right (good LED/LDR selection) it can do that trick.

[Kipper4]

Yer that might be critical Sam
I found some stuff amongst it this

https://masutopedals.wordpress.com/category/magnavibe/

Fuzz dogs pedal parts uk advocates a high intensity red led and ldr with 1M dark.

[nocentelli]

On second reflection this is not what I thought it was and could probably be done with the 4ms panner. (Twin temelus lunes) I'm not hearing the pitch modulation I thought I was last night.
Feeling tricked.

I think a panner can sound a bit pitch modulated, the speakers are not going to be exactly the same distance from the mic/ears, so there will be rhythmic phase cancellation going on. There is a dual magnavibe proposed schematic over at the other place, a google search for "dual magnavibe" is probably the quickest way to find it.