Operational transconductance amplifiers: Demystified

[ExpAnonColin]

So, after playing with some CA3080's, and breaking two , here's my understanding.  I thought I would share it with you all because before getting some 3080's I had no idea what they were or how they worked but was incredibly curious because of the possibility of a replacement for an LDR.  Excuse me a head of time for any misunderstandings I may have, I'd be incredibly glad to be corrected.

An operational transconductance amplifier is basically the same as a operational amplifier, with one nifty addition: the amplifier bias input, at pin 5.  What this is is basically an input for an external current source.  What the input does is it takes the current you give it and uses it to determine the gain of the circuit.

So, let's say you have a OTA set up as an inverting amp with a gain of about 2.2 (100k R1, 220k R2).  With a standard op amp, this would basically just double the gain of your signal, but with the OTA, things get a bit more fun.  If you put a waveform into the amplifier bias input (the 3080 wants .7V or so), then at the peak of the waveform the current is greater, so the gain is too.  At the bottom, the current is lower, so the gain is as well.  What you get is a device that modulates the gain of your signal from 2 (correct me if I'm wrong) to 0.

So, the most obvious use would be a tremolo-you use whatever LFO you were driving the LED with and put it into the amplifier bias input, keeping in mind that you want about .7V.  Whenever the waveform goes down, so does your signal, and vice versa.  The big advantage, though, is that with an OTA you can go a LOT faster, hence, ring modulation.  The frobnicator ( http://geofex.com/PCB_layouts/Layouts/frobn.pdf ) uses a 3080 in such a manner, as do the modular synth ring mods made by Serge.

Does this make sense/am I correct?  Any important stuff that needs adding?

-Colin

[Tim Escobedo]

You need to do some serious reading. You got the general idea. The easiest to understand tutorial on OTAs I recall was serialized in Nuts and Volts several months ago by Ray Marston. He has a book or two out there (usually out of print) with the info, too. Covering some of the basics in a down to earth manner. The hardest part is that there are relatively few OTAs out there, so you may have to find info specific to 3080, 13600, 3280, etc.

The datasheet for the 13600/13700 from National is pretty good, too, if a little more obtuse. Lots of examples in that one.

There are a few interesting tutorials floating around the net, too. The synth DIY sites generally have more info on these things than stompbox sites.

[Paul Perry (Frostwave)]

www.national.com/ds/LM/LM13700.pdf
ya just beat me to it! the LM13700 datasheet is pretty good... the best practical advice I can gice, is to be VERY careful never to run more than 2ma into the bias pin of an OTA. Instant death if you do. I always hang a 10k resistor on the bias pins FIRST, before I do anything else.

[Boofhead]

OTA = Operational Transconductance Amplifier

A transconductance amplifier means you have an voltage input and a current output.  The gain of such an amplifier is the output current divided by the input voltage. The ratio of a current to a voltage is termed conductance, hence the name transconductance.  In most OTA the gain is controllable via another pin in fact the gain is usually proportional to the current into the control pin.  A transistor can be considered a transconductance amplifier - that's one reason transistors are the basis of an OTAs.

A normal opamp is a voltage amplifier, voltage input and voltage output.  The gain of an opamp is the ratio of output voltage to input voltage (at the opamp + and - terminals).  The opamp gain is fixed. Opamps have a very large voltage gain (in the thousands), in fact it's too high for most applications, so you use feedback to set the gain to a lower more useable value.   With feedback the opamp gain itself isn't changed it's still high, what happens is the feedback *connection*, consisting or external resistors results in a lower gain for the combined opamp + resistor circuit.  When you have gain pots to set the gain you are changing the external feedback components which set the gain not the opamp gain itself.

OTA operate more "open" loop and that's why the gain control pin can control the gain as it does.

[Paul Perry (Frostwave)]

OTAs REmystified! maybe someone here can benefit from these links.. at least there is the 3080 datasheet! (afraid it is all a bit much for me.. :oops: )


http://et.nmsu.edu/~etti/winter98/electronics/grise/wrg.html

http://homepages.compuserve.de/Stromeko/diy/OTA.pdf.

http://www.uni-bonn.de/~uzs159/ota3080.html

http://www.intersil.com/data/FN/FN475.pdf

[keninverse]

Can anyone describe how to use an OTA as a variable resistor upon inputting voltage...

[puretube]

read all about it in above mentioned:

http://www.national.com/ds/LM/LM13700.pdf  

datasheet...

[aron]

archive

[Brian Marshall]

What are some unique uses of these in guitar effects?

circuits on the datasheet look really bulky for not doing much.

[Jay Doyle]

The CA3080 OTA is at the heart of the Ross Compressor and Dynacomp.

They can be used for VCAs, voltage controlled resistances, voltage controlled frequency stages, phasers.

Very useful actually, just ask Bob Moog  

[puretube]

OTA = Small Stone;
OTA = Oberheim synths;

(sorry, I know this is a little generally speaking...(just to mention the importance!)

[ExpAnonColin]

Thanks for the responses, guys.  I believe the Moog Ring Mod uses a LM13700.

-Colin

[Brian Marshall]

so it's just current gain instead of voltage gain???  Sorry i dont feel like reading those data sheets, besides I think you want to show off your Knowlege.

I have a prety good understanding of 'normal' opamps.  I never even really paid attention to these ones.

[Boofhead]

current gain instead of voltage gain

Perhaps more correct is current output instead of voltage output.  The input is voltage in both cases.  When you have voltage input and current output the gain is how much current out for how much voltage in - the gain is techinally called transconductance.   The term "current gain" is more for something with current in and current out (transistors can be viewed this way) - that would be a current amplifier.

[Paul Perry (Frostwave)]

The input to the OTA is in fact current. How much current, depends (in most circuits) on the voltage applied to the relatively high resistor connected to the input. So you could think of it as a voltage input if you wanted, I suppose.
The output is a current as well, but having a load resistor for the current to run through develops a voltage. So I suppose you could think of there being an output voltage too, if you like. But note , that the output voltage of a normal op amp is (within reasonable limits) independant of output load, the OTA output voltage is directly proportional to it.
Likewise, the gain control is via a current, which (often) is supplied by a resistor with avariable  voltage input. So again you can think of it as having a control voltage at the other end of the resistor, if you like. But, everything goes a bit pear-shaped at the low current extremities, where the input resistance of the OTA, neglected so far, starts to look too big. And you find (for example ) that the bottom end extremity of your LFO is behaving very badly..
Incidentally, I can't believe Moog would have been so cheap to make a ring modulator from an LM13600.. has anyone had a look inside, and if so, is there a noise gate in there? because, a LM13600 does not make a quite ring modulator.. if it did, I'd be quite a few $$ better off  :?

[uncle boko]

When I was working on the accounts at EH UK back in 1980 or so, I remember seeing two or three sheets of information and tricks on this OTA. Pity I didn't take a copy when I looked through the circuits draw! Just thought I'd tell you that - bloody annoying really.

[Boofhead]

The input to the OTA is in fact current. How much current, depends (in most circuits) on the voltage applied to the relatively high resistor connected to the input. So you could think of it as a voltage input if you wanted, I suppose.

Paul, technically I don't quite agree with that view.   The main reason is OTA implies transconductance which implies voltage input.  With a BJT based OTA the transconductance (gain) in terms of input voltage input is very predictable and well defined - basically determined by the gm of the BJT's at the input differential pair inside the OTA.  The current gain on the other hand is dependent on the gains of the transistors and this isn't well defined at all.   A BJT can be views as either a transconductance or current amplfier, so your view isn't wrong it's just not the intended view and understanding circuits with that view makes life difficult.

The other thing is the input resistors aren't in fact high value series resistors.   The input circuits usually have low value resistor to ground as well, so in fact the input circuits are *voltage dividers*.  The whole idea of the voltage divider is to ensure the input voltage swings do not exceed the saturation voltage of the differential input stage (which is fairly well defined too).  The predistortion diodes form kind of non-linear divider to linearize the differential input stage.   Usually you try to keep the grounded resistors low value so the loading effects of the input resistance of the diff-amp doesn't affect the divider.

[Paul Perry (Frostwave)]

The other thing is the input resistors aren't in fact high value series resistors.   The input circuits usually have low value resistor to ground as well, so in fact the input circuits are *voltage dividers*.  

Mate, you are entirely correct. Thanks for pointing that out!

[Jason Stout]

Brian Marshall Wrote

besides I think you want to show off your Knowlege.

Some call it sharing.

That is what this forum is about.

[TELEFUNKON]

bump for the april 2010 24dB filterers