Using opamp saturation to limit voltage

[Alex Nikitin]

I've looked carefully on the data for the AS3360. The maximum gain is achieved at the control voltage equal to the reference voltage, 1.7V, so there is 0.8V to spare on top of it, and the lowest control voltage is actually at 0V so there is no problem whatsoever to use just a plain diode to limit the negative control voltage and a transistor to limit the positive (to reduce current required from the reference pin). 

Cheers

Alex

[Kleinstein]

How bad a OP amp reacts to clipping at the output depends. Especially some rail to rail types can recover relatively fast from clipping and could be used limiting. This goes so far that I saw the suggestion to use it for a precision full wave rectifier.

[magic]

According to LT Spice the circuit is stable with 100K on the input and 100nF on the output  . Caps around Q3/Q6 don't work that well (I've tried in the sim).

Interesting, you're right. I forgot that Miller comp falls apart when 2nd stage gm decreases too much without corresponding loss of 1st stage gm. High source impedance makes this happen by unloading the output transistor...

BTW, for minimum offset voltage the collector load resistor should be chosen to conduct half diff pair current near the threshold. So maybe 1k, not 6k2.

I've looked carefully on the data for the AS3360.

And this is probably the sanest way out of this situation

[Alex Nikitin]

BTW, for minimum offset voltage the collector load resistor should be chosen to conduct half diff pair current near the threshold. So maybe 1k, not 6k2.

1K there will noticeably restrict the voltage ranges, both for the supply and the limiting levels. Ideally one should use a 2-transistors current source in place of 10K resistor, that will allow to balance the differential properly and provide a nice temperature compensation as well. Gets somewhat complicated though.

Cheers

Alex

[level6]

I've looked carefully on the data for the AS3360. The maximum gain is achieved at the control voltage equal to the reference voltage, 1.7V, so there is 0.8V to spare on top of it, and the lowest control voltage is actually at 0V so there is no problem whatsoever to use just a plain diode to limit the negative control voltage and a transistor to limit the positive (to reduce current required from the reference pin). 

Cheers

Alex

This work very well. Thank you!

I will need to scale the input, though, so that +5V scales to +1.7V. The expected external control voltage range is 0-5V, but anything above that, up to 12V, is possible.

[Alex Nikitin]

This work very well. Thank you!

I will need to scale the input, though, so that +5V scales to +1.7V. The expected external control voltage range is 0-5V, but anything above that, up to 12V, is possible.

Just add 5K1 resistor in parallel to C1 in my second circuit.

Cheers

Alex

[level6]

This work very well. Thank you!

I will need to scale the input, though, so that +5V scales to +1.7V. The expected external control voltage range is 0-5V, but anything above that, up to 12V, is possible.

Just add 5K1 resistor in parallel to C1 in my second circuit.

Cheers

Alex

Perfect! Thank you for this.