Troubleshooting Op Amp Attenuation in Gain-Adjustable Amplifier Circuit

[tomizzo]

I have a question regarding an amplifier circuit that I’m designing. I was given a template to improve upon, and have made several changes to the circuit. Refer to the attached circuit schematic.
Essentially, I need to create an amplifier that has a gain from 0 to 10 using an op amp that is powered by two +/- 20V voltage sources. I have a potentiometer that adjusts the gain of the input, and then have the op amp configured to have a maximum of a 10 gain. This is set by the 9K ohm and 1K ohm voltage divider. However, I am noticing attenuation of the output signal from what I’m expecting.
The op amp should be able to provide an output of +/- 20V. When I set the input signal to be 2 volt peak and have the gain configured to 10, I should expect an output to be 20v peak. However, looking at the output signal, (green trace, channel b) the output only reaches 16.279V. For some reason the signal is attenuated as if the op amp is railed. I can’t explain why this is happening.
Any ideas?

 

[analogdesign]

A couple of things come to mind. First those diodes will limit the swing to two diode drops below VCC (minus whatever the drop from the series resistor is). Second, is it an ideal opamp? Most opamps can't swing all the way up to the rail unless they are specifically designed to do so.

 

[Averagesupernova]

It also looks like the signal on the scope shows a fairly constant slope on the waveform which to me indicates you are pushing the max slew rate.

 

[tomizzo]

A couple of things come to mind. First those diodes will limit the swing to two diode drops below VCC (minus whatever the drop from the series resistor is). Second, is it an ideal opamp? Most opamps can't swing all the way up to the rail unless they are specifically designed to do so.

Well the thing I don't understand is when I increase Vcc to +/- 40V, the op amp output voltage still doesn't increase. It's almost as if it's stuck at 14V maximum output. Something has to be limiting it but I just can't figure out what.

 

[tomizzo]

Also, I just found that the op-amp has no difficulty reaching the rail voltage from the power supply, there must be something in the circuit that is limiting the output voltage to 14V, but I can't seem to put my finger on it.

 

[analogdesign]

Also, I just found that the op-amp has no difficulty reaching the rail voltage from the power supply, there must be something in the circuit that is limiting the output voltage to 14V, but I can't seem to put my finger on it.

I think the Darlington devices (Q1 and Q2) are holding the voltage. Their bases are constant so increasing the supply just increases the current through the series resistors. How is the voltage at the emitter of the darlingtons set?

 

[Integral]

To operate correctly OP amps need feedback. A OP amp drives the output up as long as the there is a difference in potential across the inputs. The output of your OP amp has no effect on the input, therefore it is driving to the rails.

Sitting OP amp gain is done with resistors feeding back the output voltage to the - input.

This is well known electronics, you should be able to find circuits online.

 

[Averagesupernova]

The output of your OP amp has no effect on the input, therefore it is driving to the rails.

There is feedback unless I am reading the schematic wrong.

 

[Baluncore]

The four diodes will have a greater voltage than the two Darlington Vbe total.
The current through the Darlingtons is not limited. Try removing at least one diode.

Also put a resistor of a few ohms in series with each emitter to limit quiescent current.
That resistance is compensated for by the Darlington beta and the feedback loop.

 

[analogdesign]

To operate correctly OP amps need feedback. A OP amp drives the output up as long as the there is a difference in potential across the inputs. The output of your OP amp has no effect on the input, therefore it is driving to the rails.

Sitting OP amp gain is done with resistors feeding back the output voltage to the - input.

This is well known electronics, you should be able to find circuits online.

As averagesupernova said, there is feedback in this circuit. The output (which is buffered by the Darlington devices) is fed back to the non-inverting input of the op amp through the voltage divider formed by R4 and R5.

 

[AlephZero]

The output (which is buffered by the Darlington devices) is fed back to the non-inverting input of the op amp through the voltage divider formed by R4 and R5.

... and that is the right thing to do, to compensate for the nonlinear behavior of the circuit when the output goes through zero. Taking the feedback from the op-amp output would give severe crossover distortion.

 

[meBigGuy]

The circuit has a gain on 10 from the plus input set by R4/R5

Possibly both transistors are turned on because of mismatch between 1N914 and Vbe. I can't tell what colors match what points and can't see the blue at all. Look at both transistor currents and scale the trace so it isn't going off screen.

 

[Baluncore]

The MC33074L is only rated to 44V total supply. Output current is limited to 10 mA, that is insufficient to pull the 1k3 Darlington base resistors more than about 15V from either rail.

 

[meBigGuy]

Except he said the opamp output swung (swang? swinged?) the entire range. Unless he did that unloaded.

 

[Baluncore]

This circuit uses current mirrors to provide a fixed 1 mA base current to the op-amp output diodes.
The 470R limits op-amp output current by turning on the Darlingtons only when needed to boost the output current.