Why does OTA sine wave oscillator need no amplitude control?

Discussion in 'General Electronics Chat' started by Veracohr, Feb 18, 2016.

  1. Veracohr

    Thread Starter Well-Known Member

    Jan 3, 2011
    Attached is a sine wave oscillator schematic from the LM13700 OTA datasheet. Does anyone know why it doesn't require amplitude stabilization like Wien bridge or phase-shift oscillators do? I built it up and verified that the amplitude stays steady, it doesn't grow to the device's saturation limits. The amplitude is set by A1 and the 30 ohm resistor.

    Screen Shot 2016-02-18 at 4.57.23 PM.png
  2. Lestraveled

    Well-Known Member

    May 19, 2014
    First, I want to say that it is cool that you are using OTAs, because these devices can do functions that standard op-amps simply can not do.

    OK, to your question. This circuit is affected by temperature, so if you hit it with a heat gun you will see an amplitude and frequency change. As to why it does not self saturate is that A1 is set up as a comparitor and that it changes state as the output crosses zero. The time constants and the gains of the stages keep it from saturating after that. So, It will be hard to make this oscillator self saturate.

    In other words, A1, the zero crossing detector, starts the same process every time the output crosses zero, which keeps the everything in control.
    Last edited: Feb 18, 2016
  3. AnalogKid

    AAC Fanatic!

    Aug 1, 2013
    A slightly left-handed way to view this circuit is that it is a square wave oscillator, which like all square wave oscillators makes a constant amplitude output regardless of its frequency, followed by a 3-stage lowpass filter. The output of A1 is a square wave. The outputs of each succeeding stage are more and more filtered, rounded, and sine-looking. Not exactly what is going on, but close enough to explain why there is no light bulb.

  4. hp1729

    Well-Known Member

    Nov 23, 2015
    Interesting. Is this audio? I have to try this one. Attached is a version with pin numbers for those who may want to also..
  5. Veracohr

    Thread Starter Well-Known Member

    Jan 3, 2011
    The output of A1 is actually pretty sinusoidal on the circuit I built, but I can see what you're saying. I was thinking of it as more of a voltage-controlled phase shift oscillator than a filtered square. The simulation I ran shows the output of A1 somewhat more square than my actual circuit, though still pretty rounded.

    It does change a little with temperature, but not hugely. Except when I was supplying the bias current from a current source that WAS hugely affected by temperature. When using a voltage source with resistors like it shows in the schematic, I got a frequency change of 0.068% per degree C.

    Yes I've been testing this at audio frequencies, although the datasheet says the circuit can operate up to 50kHz. I'm sure it could go higher with smaller capacitors. It lists an open loop bandwidth of 2MHz.

    In my circuit I actually used TL072s as buffers instead of the darlingtons included in the LM13700.

    I got interested in OTAs when I started looking into analog synth designs, especially voltage controlled filters. Some classic synths of the past used OTA-based filters, like certain versions of the Korg MS-20. For my senior project I made a MIDI-controlled analog effect processor with three OTA-based effects. At the moment I'm running various tests of oscillator circuits to write a technical comparison paper. For a writing class actually, not an engineering class.