0
\$\begingroup\$

I've constructed a basic dual op-amp LFO as in the below circuit

schematic

simulate this circuit – Schematic created using CircuitLab

The variable resistor R5 is used to alter the rate of the LFO, I'm unsure of the exact frequencies, but this is used in an audio application to control a VCA. It runs on a dual rail power supply at +/-9V, and uses 0V as VR. It works fine.

What I would like to do is control the rate of the LFO (currently done by R5) with another, similar LFO running at a different speed, so the overall rate of the LFO is not constant, and some randomness is added, as if a potentiometer were being constantly adjusted.

My question is - assuming I had another LFO as above, how would I use that to control the rate of the first LFO? My first thought was to use a standard BJT transistor, but I am unsure how to implement this. Can anyone give some pointers?

\$\endgroup\$
  • \$\begingroup\$ Please define LFO \$\endgroup\$ – Scott Seidman Sep 12 '18 at 16:11
  • \$\begingroup\$ ... and VCA, while you're at it. \$\endgroup\$ – Scott Seidman Sep 12 '18 at 16:13
  • \$\begingroup\$ I understand both abbreviations but not the specs. F1 range, F2 slew rate, F2 range of random slew rate, F1 Amplitude control where say F1 is audio out and F2 is control of VCO and VCA is for which F1 or F2 amplitude warble or amplitude audio? \$\endgroup\$ – Sunnyskyguy EE75 Sep 12 '18 at 16:14
  • \$\begingroup\$ low freq osc & v ctrl amplitude.. gives some good sound bytes for specs instead \$\endgroup\$ – Sunnyskyguy EE75 Sep 12 '18 at 16:18
  • 1
    \$\begingroup\$ I know the abbreviations, too, but many won't. I suspect VCA is voltage controlled amplifier. \$\endgroup\$ – Scott Seidman Sep 12 '18 at 17:20
1
\$\begingroup\$

Your design is a simple phase shift oscillator that uses series R to control loop gain.

Your requirements have 2 main parts; a triangle VCO or Pot control LFO, that has warble or jitter on the variable input included. I will address part 1 using a transistor but you can use a Pot instead.

The latter must be BETTER defined more in terms of random frequency range and modulation or by sound byte example.


part 1.

Perhaps a better way is a simple Relaxation Oscillator to use shunt feedback to reduce amplitude at the virtual ground point so that voltage drop is a virtual AC null. (this is a key point)
The series R regulation induces distortion from voltage drop at max R.

[My Example Simulation] enter image description here I answered this already for a sine wave 10MHz VCO with a 4 or 5:1 tuning range which is a bit harder.

However, as Voltage bias decreases, impedance rises and voltage rectification occurs making the sine wave look somewhat like a full wave rectified at low frequency when R is limited by a shunt resistor. The series R limits the VCO for max f.

Until you have design specs you cannot design anything well;
F: max-min, tuning ratio, VCO gain [MHz/V or kHz/V] VCO linearity, Vin range
V: amplitude Vpp and DC offset and amplitude variation tolerance,
slew rate on amplitude changes vs df/dt,
THD: Max, Nominal sine .. or just a triangle wave

\$\endgroup\$
0
\$\begingroup\$

My first thought was to use a standard BJT transistor, but I am unsure how to implement this. Can anyone give some pointers?

Remember that if you have a BJT, you also need to deal with the ~0.7V drop. There is also current from the base that (Ib) is added to the emitter current (both effects make it much less like a current controlled resistor)

A better way is to use a mosfet or JFET to make a voltage controlled resistor.

enter image description here Source: http://www.nutsvolts.com/magazine/article/fet_principles_and_circuits_part_1

This presents its own challenges as FET have variable transconductance (gain) curves so getting the same result on one FET part vs another can be difficult. That means the same amount of voltage will not give you the same resistance. FETs usually requires calibration or feedback to make them usable (or just use as a switch).

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.