Snippet: continuously variable gain/polarity

[earthtonesaudio]

I saw something like this on the web ages ago, now I can't find it, so I decided to draw it to the best of my memory, hopefully someone will correct my errors, if any.



Essentially you have a very simple circuit which adjusts gain, continuously, from -1 to zero to +1.  Most applications would require a buffer before this circuit for acceptable performance, but this would probably be something you'd integrate into a larger circuit anyway.  I drew this as a "textbook" op-amp circuit, with ground referenced in/out and bipolar power supply implied.

I think this would be a nice thing to put in active splitter/looper projects, or in anything with an effects loop.

[Taylor]

So this could be used in the feedback loop of a phaser to get either notches or peaks? If so, very cool, and it's going into my 12-stage Phase 90.

[Thomeeque]

Essentially you have a very simple circuit which adjusts gain, continuously, from -1 to zero to +1...

Hi! I did not get it fully yet, but I'm almost 100% sure, that it would go from -2 (positive input on ground, R1 makes only ballast for input signal, Gain = -R3/R2)..

[earthtonesaudio]

Ah, you're right Thom!     Looking at it some more, I think it works out if all resistors are equal.

I wish I could find that original schematic...

[Lurco]

page 8: IC1D 

[Thomeeque]

page 8: IC1D 

I believe that's it - hmm, very smart! 

Just note, that it requires very low output impedance of input signal source (like direct ouput of opamp), so the gain is exactly (or very close to) +2 (1 + R21/R20) from the non-inverting input's point of view.

T.

[Thomeeque]

 So, to make this work:



Just set R₂=R₃ (R₁* value is not relevant for the basic concept**) and state, that R2 is much bigger than Z_out of the preceding stage (input signal source).

Edit: Equivalent of Craig's R14 should be there too (for stability reasons), even it is not relevant for basic concept as well.

T.

* It just should be much bigger than Z_out of the preceding stage as well of course.
** final gain = -R₃/R₂ + (1 + R₃/R₂)*x, where x represents R1 pot divider ratio (0 at min, 1 at max)

[earthtonesaudio]

Excellent!  I'll fix the drawing sometime.  Glad to know I wasn't just imagining things... 

[earthtonesaudio]

Fixed:

[Thomeeque]

Fixed:

Besides the equation below. IMO there should be something like R2 = R3, R4(5k) = R2 || R3, R1 value is not related to the rest..

T.

[earthtonesaudio]

Thanks Thom.  Fixed now, I hope:

[Processaurus]

Hi, thanks for reminding us of this circuit.  This would be good for parallel effects loop circuits like the splitter-blend.  

Note that the input impedance of the + input is quite high, so it can't do anything to the signal, and R4 isn't doing anything by reducing current, and can be omitted, with the pot's wiper connecting directly to the + input.

[edit] Thom, I just read what you said about R14 being there for stability.  It looks like it is there so that with the vibrato/phase switch on, the output of the previous stage won't be loaded by the pot, if it's all the way down to ground.

[R.G.]

R4 is there for DC balance reasons for the opamp. In opamps which take some appreciable bias current, the opamp will produce a more accurate DC output if both inputs have equal resistances to the bias voltage source. That forces the bias currents to come through equal resistances and reduces the bias-created offset which is then multiplied by the DC gain. For AC circuits, it's a nicety, but not necessary.

Notice that, like most opamp application notes, this circuit assumes a bipolar power supply. Running it from a single 9V supply will require more circuitry.

I first ran into this circuit in the National Semiconductor opamp compendium application note. It's far older than that.

[earthtonesaudio]

Yeah, I think it would be nice for spltter/blend circuits because you can combine your gain and polarity controls into one knob.  I wish I could find an appnote where this appears to know the "real name" of this circuit.  That knowledge probably would have saved me dozens of minutes on Google.   

[R.G.]

It's called a "switch hitter".
See http://users.ece.gatech.edu/~mleach/ece3050/sp04/OpAmps01.pdf.

[swt]

i've been using this as an inverter for the resonance in my phasers, it will give you positive and negative feedback...take a look at the infinit phase, from 4mpedals, and paul perry also use something similar in his ms20 filter clone. it can be done with 9v also, by using a vr. it's a useful snippet, that mark brought us a long time ago...

[earthtonesaudio]

Ya know, I never in a million years would have thought to look for this under "switch hitter."  Huh.