Check My Work: Op-amp Circuit

Isn't the object to have the output follow the input?

The LM358 can have inputs go only to VCC-1.5, then you lose feedback effects. ...Jim Thompson

Whoops. I knew there was a good reason to ask! I took care of the _output_ not going to the rail, but not the _input_.

I need to think about this -- I think I can fix this in one of three ways:

With the rail-rail amp I'd use the same topology, just jigger the resistor values to get the same effect.

With OpAmp rails greater than needed swing, then clamp...

But a rail-to-rail OpAmp (input AND output) should need nothing else if powered from same rails as ADC. (Most everything I know of can stand 0.3V over and under rails without any problems at all.) ...Jim Thompson

What's the active range of the ADC?

John

The way I have it wired, pretty much 0.3V to 4.7V. It's a 16-bit ADC in an application that could get by with less.

Why not just use a r-r opamp powered from ground and +5?

John

How about an emitter follower with +5V Vcc?

The way I would do it, after looking at a couple of older part spec's, is to use something like a LM393 comparator. In the saturation state, it drops to 250mv, I don't think that is much of an issue here. The pull up R can then be clamped with a 4.7V zener. The output looped back to the inverted input. This gives you only 2 added components in the mix..

I did something like this years ago where the input could get up to

I supplied the comparator with 15 volts and clamped the pull up R with a 5.1 Zener. This simple circuit allowed me to have voltages on the input up to 15 volts with out fear of damage but still operated as I needed it on the output..

Just a thought, you may have different ideas and different components to use, I'm sure..

Jamie

I think (3) is the winner; a two-resistor attenuator on the input will also make open inputs float to LOW in a reliable fashion, which is sometimes a very good thing. The feedback, then, comes from a center-tap on R3, but then there's no need for R4 at all. So, it's just adding two resistors, and using some of the op amp gain (which you've already paid for).

Isn't there a delay before one of the clamp opamps comes out of saturation? I remember delays for opamps coming out of saturation but it was long ago (decades) when I was doing analog design.

Cannot remember how long LM358 or LM324 opamps take to do that.

Or, is the coming out of saturation delay short enough to not worry about, I'm wondering if there would be spikes on the clamped output while the clamping opamp got it act together.

But I can't see how to keep the opamp in feedback mode easily without loading the input, with your circuit.

Maybe I shouldn't worry about that delay?

Grant.

I wouldn't worry about the OpAmp recovery unless you have a very fast signal. If you do, there are more exotic configurations that keep the OpAmp out of saturation. ...Jim Thompson

Sure, something like a bakers clamp.

Jamie

in

A R-R buffer with anti-parallel schottkies in the output would guarantee something pretty close to Tim's spec.