Op Amp Circuit Gain Does not Match LTSpice Simulation

[epsolutions]

I would prefer DC coupled because there is a need to preserve the wave shape at low frequencies. But I am open to suggestions, if the DC coupling is what is limiting/clipping gain at output.

The input is approx line level audio. It is a bipolar signal with a +/-1.5V swing.

Many thanks for your input so far.

 

[danadak]

Is this more what you want ?



Regards, Dana.

 

[epsolutions]

Thank you for sending that circuit. This is how LTSpice simulates it. Since there seems to be a recurring discrepancy between different sims, and sims and real life, I will not add further to this thread but rather research elsewhere and get back if and when I can provide some useful information.

 

[Harald Kapp]

why does the sim (.asc file attached) show an output of twice of what I am measuring in the constructed circuit?

But how is this related to the offset, or some other factor?

The simulation uses a model of the real circuit (opamp). Any model makes simplifications to speed up simulation time. Example: set the supply voltage V1 to 9 V:

The simulation shows an output voltage of 8 V whereas in a real circuit the output voltage of the opamp is limited to ~ 7 V (V1 - 2V), cf. post #13 or see the datasheet. Subtract 2 × B_BE for the base-emitter juctions of Q1 and Q2 the output at the emitter of Q2 can be max. ~ 5.8 V.
Or in other terms for the output voltage to reach 9 V, the opamp's output voltage needs to be at least 10.2 V. Add the top margin of ~ 2 V ( Vsupply - Vopamp) and the supply voltage needs to be at least 12.2 V (minimum) or 13.2 V when assuming a 3 V margin (again, see the datssheet).

The simulation also shows the voltage at the output of the opamp to go to -100 kV. You do agree that this is unrealistic, do you?

You need to be aware of the limitations of a simulation. Any simulation needs to be taken with a grain of salt, especially when the real circuit shows a different behavior. Looking not only at the obvious points of interest (Vout) but inspecting other nodes as well helps to identify potential issues (Vopamp).

 

[danadak]

Using a R-R OpAmp we get some better results :



Choice of OpAmp dictated by what was already in sim, eg. spice model. A general
purpose RRIO OpAmp will pretty much do. Note slew rate of OpAmp shown probably a
little light, you need something on the order of 1.25 V/uS min for 20 Khz input.

You obviously need to heat sink the 3055 output transistor.


Regards, Dana.

 

[Ylli]

Danadak, recheck your schematic. You have the feedback resistor connected to the op-amp output, it should be connected to the top of the inductor.

 

[danadak]

Thanks Ylli for catching that, here is updated schematic :



I checked the phase margin, that looks OK as well.


Regards, Dana.