RE:Output clamping 2

[fastfred]

I have a diff-opamp that runs on +/-9vdc and would like to limit the output rail at +/-5vdc.What kind of zener or other component arrrangement will work, please help fast because I need something soon.

 

[nbucska]

Describe circuit, give specs.

<nbucska@pcperipherals DOT com> subj: eng-tips
read FAQ240-1032

 

[fastfred]

I have a standard 741 IC with a negative feedback gain of 8.333,with a 2.5volt ref at the negative and 1.9 to 3.1vdc range at the positive input,and the output has to rail at +/-5vdc, the supply voltage is +/-9vdc.How can I limit the output to +/-5.5vdc max?

 

[felixc]

The brute force solution would be to have a series resistor at the output to limit the current, and to clamp with zeners, or a bidirectional avalanche diode. If you need any current capability after this you can add a unity-gain buffer.

 

[IRstuff]

Why not use comparator instead?

TTFN

 

[cbarn24050]

Hi, use 2 4.3volt zeners in series back to back across the feedback network. (pin 6 to pin 2)

 

[logbook]

Sorry but back-to-back zeners as described is a really nasty solution. Zeners get really "soggy" towards the knee and start conducting.

You should bias the zeners individually, giving them say 5mA current each. Then you can use this as a clamp point for a low leakage diode. We are talking about one zener, one bias resistor, and one low leakage clamp diode for the +5V clamp and this same circuit again for the -5V clamp. If you have +5V and -5V power rails you could of course just use schottky diodes to these power rails. You still need the series protection resistor from the opamp in all schemes.
schemes.

Note that putting zeners across the feedback network does not limit the amplifier output as desired. It drops the gain into the non-inverting input from 8.333 to 1, but that still means that increasing the non-inverting input will increase the output.

Whilst we are on the subject, your basic circuit bias is wrong. You have a non-inverting signal of +3.1V to +1.9V, that is nicely centred about the +2.5V reference level. When you put in +2.5V to the non-inverting input, given that the inverting input resistor is also at +2.5V, how much are you expecting at the output? If you think the answer is 0V then you flunked opamps 101.

Try again using nodal analysis, following the currents.

 

[BigTom2005]

Why not use another op-amp (Follower circuit)---the feedback path is simply a short circuit, so the output equals the input. Then power +Vcc with 5V, and if the positive portion of the output signal exceeds +Vcc, it will saturate at 5V(Or +Vcc). You can probably do this also with a BJT, and have the output Saturate at Vcc (Common Emitter), where the Vcc is applied to the collector side of the circuit.

The last way would to be to use a comparator.

If cost is an issue, then go with the diode approach. You might want to consider it if the bandwidth of your amplifier is large. Diodes work great at high frequencies.

Good luck.

 

[fastfred]

thanks guys for everything,my goal was to have the output travel linearly from 0 to +5v by varying the noninverting input from 2.5v to 3.1 and the output also vary from 0 to -5vdc by varying the noninverted input from 2.5v to 1.9vdc. The feedback resistors was 20k(Rf)and 2.4k(Ri) and the noninverted input voltage divider was 20k(pin-3 to ground) and 2.4K.I was able to install a 220ohm current limit in series with output and two 4.3v zeners cathode to cathode to limit the output to +/-5vdc, thanks everyone for all the help(great forum

 

[hoytclagwell]

Note that linear-tech has exactly the opamp you need---it allows you to set the output voltage limit. See LT1970