School Project Help.

[ozgunoral]

Hello,

I need a little advise for a little project.

I need to build a smart(with automatic gain control) amplifier with the following characteristic:

When the output voltage swing exceeds an external limit set by the user (Vref), amplifier decreases its gain such that the output voltage swing remains aroud Vref.

The circuit is supposed to be operating up to 20 kHz and maximum gain should be 40.
Supply voltages are +5 and -5 volts.
Min Vref=100mV and maximum deviation can be 5%.

Thank you for your help.

 

[KrisBlueNZ]

Hi again,

Have a look at the LM13700. Data sheet is here: http://www.ti.com/lit/ds/symlink/lm13700.pdf

It's a dual operational transconductance amplifier that can be used as a variable-gain amplifier. It can operate from +/- 5V. You need to monitor the peak-to-peak amplitude at the output (using a full-wave "ideal diode" circuit - Google it) and smoothing capacitor, compare it to the external voltage, and use the result to control the gain.

You'll need to be careful with the control loop response, so it responds fairly quickly to an overload but not so quickly that it overshoots. There is a delay in the response of the amplitude detector, because of the capacitor. Its response following an increasing amplitude and its response following a decreasing amplitude are set independently. You'll need to do some testing. I created some waveforms with sudden changes in amplitude and burnt them onto a CD. Then you can use a portable CD player as your test signal generator!

Edit: Actually you may not need the ideal diode. Just compare the instantaneous output voltage to the reference voltage, and a negative reference voltage derived using an op-amp, and while it exceeds the reference, gradually charge a capacitor that causes the gain to ramp down. Ramp the gain back up at a slower rate by discharging the capacitor through a higher resistance.

 

[ozgunoral]

Hi again,

Have a look at the LM13700. Data sheet is here: http://www.ti.com/lit/ds/symlink/lm13700.pdf

It's a dual operational transconductance amplifier that can be used as a variable-gain amplifier. It can operate from +/- 5V. You need to monitor the peak-to-peak amplitude at the output (using a full-wave "ideal diode" circuit - Google it) and smoothing capacitor, compare it to the external voltage, and use the result to control the gain.

You'll need to be careful with the control loop response, so it responds fairly quickly to an overload but not so quickly that it overshoots. There is a delay in the response of the amplitude detector, because of the capacitor. Its response following an increasing amplitude and its response following a decreasing amplitude are set independently. You'll need to do some testing. I created some waveforms with sudden changes in amplitude and burnt them onto a CD. Then you can use a portable CD player as your test signal generator!

Edit: Actually you may not need the ideal diode. Just compare the instantaneous output voltage to the reference voltage, and a negative reference voltage derived using an op-amp, and while it exceeds the reference, gradually charge a capacitor that causes the gain to ramp down. Ramp the gain back up at a slower rate by discharging the capacitor through a higher resistance.

Thank you I've saved the document and your response to a folder and shared with my partner.

There are also bonus points for designs that don't have opamps but transistors only. Is there any chance that you have any ideas?

And I think yours is the perfect solution but we are expected to make more primitive designs so that we can show our knowledge of the basic components.

 

[KrisBlueNZ]

Well, you can use a depletion-mode JFET as a variable resistance to provide adjustable attenuation, or you can use an LDR with an LED. And you can make an op-amp or a comparator from a long-tailed pair.

 

[BobK]

If I remember correctly, a long tailed pair with an variable current source on the emitters works as a variable gain amplifier.

Bob