0
\$\begingroup\$

enter image description here

Can any one help me out to design a circuit with opAmp using this transfer curve?

Here, OpAmps +V(sat) = 15v and -V(sat) = -15v

\$\endgroup\$
2
  • \$\begingroup\$ Are you wanting gain breakpoints as per your diagram or a fluid transition from a gain of 2.5 to a gain of 1.25? \$\endgroup\$
    – Andy aka
    Feb 15, 2018 at 11:08
  • \$\begingroup\$ i want breakpoints as per my diagram \$\endgroup\$
    – habla
    Feb 15, 2018 at 14:19

2 Answers 2

4
\$\begingroup\$

Your cuve kinda looks like ań inverted tanh function.

That's the response a overdriven differential amplifier built from ordinary transistors would generate.

circuit diagram

Explanation from input to output:

The voltage divider built from R5 and R6 brings your input signal down into a range which overdrives the differential amplifier built from Q1 and Q2. As you can see we bring down the signal a lot. That's because the Q1/Q2 differential amp has a lot of gain to begin with.

Since we have only one signal we apply it to the base of Q1. The other input of the differential amplifier, the base of Q2 is grounded via R2.

We now see the following signals at the collectors of Q1 and Q2: enter image description here

That's already close to what we want. We could pick the blue signal, amplify and level shift it. That would require quite a bit of calibration and tweaking to get working though.

It's easier to take the difference of the two signals. That's what the Opamp U1 and the surrounding resistors do. It's a very simple differential amplifier with a tiny bit of gain.

The resulting difference output looks like this:

enter image description here

That's pretty close to the waveform of your drawing. Note that I've used an ideal opamp model for simulation so the output waveform slightly goes above and below your supply voltage. That'll go away if you use a real part.

Some guidelines for tweaking:

Increase/Decrease R6 to control the amount of distortion you want. The higher you go, the more linear your signal will get.

The voltage dividers R9/R10 and R8/R7 set the gain. Their ratios should be identical.

If you want to built this circuit in practice:

  • all resistors should be 1% tolerance or better.
  • U1 should be an opamp with rail-to-rail output capability
  • Q1 and Q2 should ideally be matched transistors. It will work without if you can live with a bit of asymetry though. The actual NPN transistor does not matter much. Any jellybean part will do.
\$\endgroup\$
1
\$\begingroup\$

i want breakpoints as per my diagram – habla

The basic gain should be 2.5 to coincide with the points 4,10 and -4,-10 on your diagram. I'd use a non-inverting op-amp for this with a feedback:input resistor ratio of 1.5 to acheive this.

You can then use back-to-back 10 volt zener diodes and a series resistor to reduce the gain at output voltages in excess of 10 volts. This network goes across the feedback resistor like this: -

enter image description here

It won't make perfect sharp breakpoints but it will be close. The feedback resistor needs to be 300 ohms. If you have a circuit simulator I suggest you give it a try.

\$\endgroup\$
1
  • \$\begingroup\$ @habla did you use zener diodes or something else in your test circuit? \$\endgroup\$
    – Andy aka
    Feb 15, 2018 at 19:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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