Hi everyone,

I've built a simple peak detector circuit (image attached), and in addition to my main question, I have just a general theory question I was hoping someone could shed some light on: should the ground of the input, output, and that capacitor be the same as the power supply? Or should they be separate?

(This is just a general theory question, but out of interest: my input is a series of positive-only pulses, but they tend to dip a bit into the negative... not by much, but enough to exceed the negative threshold of the op-amp. That's why I have that high-pass stage, and I was HOPING that by separating the signal and power supply ground, I wouldn't need that stage anymore. Tried it, and didn't work, but it got me curious!)

 

Most opamps using that circuit need a positive and negative dual-polarity power supply. A few "single-supply" opamps like the very slow LM358 and LM324 will work with all grounds connected to the negative power supply pin of the opamp.

A much better and faster single-supply opamp is the MC34071 single, MC34072 dual or MC34074 quad.
An input of the MC3407x opamp must never go more negative than its negative power supply pin. Your input high-pass filter does not help much.

 

Thanks for the reply, Audioguru. From a general theory point of view, when dealing with single supply op-amps do we always link up the signal and power supply grounds? That's what I'd always done, but never really gave it much thought...

 

An ordinary opamp frequently (but not always) needs a dual-polarity supply because its inputs do not work anywhere near (within about 2V or 3V) the negative supply pin's voltage and the output does not go near (within about 1.5V) the negative supply pin's voltage.

A feature that a single-supply opamp has is that the inputs work all the way down to the negative supply pin's voltage and the output goes down to very close to the negative supply pin's voltage.

A single-supply opamp can also have a dual-polarity supply if you want.

 

................. From a general theory point of view, when dealing with single supply op-amps do we always link up the signal and power supply grounds? That's what I'd always done, but never really gave it much thought...

Any electronic circuit, not just op amps, need the power ground, whether single or dual, tied to the signal ground (except for specifically designed isolation circuits). Any signal current is indirectly provided by the power supply(s) and thus must have a path to return to the power ground.

 

OK, thanks for that crutschow!

 

AC coupled peak detectors are a slave to the input signal being periodic. It is best to DC couple random pulses.