opamp detector "diode"

Thread Starter

AnalogKid

Joined Aug 1, 2013
10,989
I'm probably a few decades late to the party with this, but I had a thought. For a classic crystal radio, couldn't you replace the detector diode with one-half of an LM393 with one input grounded? I know this would require a power source, which some consider blasphemy, but the detection threshold voltage would be zero. Is this the perfect detector, or am I missing something.

ak
 

crutschow

Joined Mar 14, 2008
34,285
That's a curious question.
A LM393 is a comparator so how do you expect for it to act as a rectifier and recover the RF modulation envelope? :confused:
Are you thinking of carefully biasing it at the active point and using its linear gain to amplify only half of the waveform?
 

Thread Starter

AnalogKid

Joined Aug 1, 2013
10,989
Nope. A standard crystal radio is a half-wave rectifier. What comes out of the tuned circuit is symmetrical about ground, and the diode conducts only the positive half cycles into the peak-detector filter that recovers the audio. That detection happens only when the modulated carrier is above 0.6 (silicon), 0.3 (germanium) or 0.2 (Shottkey) volts. With a comparator powered from 6 or 9 volts, with the negative rail at zero volts, the same potential as the reference potential for the tuned circuit output, and an input stage that functions at or below the negative rail, then the output ... hmmm ... I guess it has to be an opamp rather than a comparator so it preserves the AM. OK, 1/2 of an LM358 set to a fixed gain, with the inputs referenced to the negative rail. Depending on how fast the opamp recovers from a negative input overload, wouldn't this get you some kind of envelope detection? Or at least half-wave rectification without any "Vf" signal loss? Might have to follow the 358 output with a peak detector... Still thinking this through...

ak
 

BR-549

Joined Sep 22, 2013
4,928
I don't think it will work, although it would be nice if it did.

Amplifying one half of the signal, is not the same as rectification.

We need to cut or separate the signal in half. Doing so at zero crossover would be handy.
 

BR-549

Joined Sep 22, 2013
4,928
What do you fellas get when you simulate it?

Can you set the voltage drop parameter to anything greater than zero and simulate? I am referring to a diode detector on this question. I'm not familiar with your simulators.
 
Last edited:

BR-549

Joined Sep 22, 2013
4,928
In any case, you would have to make sure all dc components are removed or nulled before entering detector.

This might be hard to do. Some dc components could vary with the signal.

Maybe a very small amount of offset might be preferable.
 
Top