Envelope detector woes

[Amish_Fighter_Pilot]

Hello. I am trying to make an envelope detector that will demodulate my signal and I have sort-of succeeded but I need help understanding what I've done(and help improving it hopefully).

So I came up with this circuit(I know it's a bit messy)



What I don't really understand is that when I feed it a simple 1.2KHz signal it gives me pure garbage, but when I feed it the modulated signal then it outputs something that makes way more sense(but loses a lot of signal strength). In the picture below: If I feed it signal A I roughly get the 200Hz signal I am going for, but I give it signal B and i get 1.2 and 2.4KHz on top of each other. I'm probably just failing to understand something really basic here so please be kind! Thanks

[Amish_Fighter_Pilot]

I have made an updated version of this circuit that is much simpler and better. That said: it still has problems. I can get the 200Hz wave out of the carrier, but only at these exact values. Trying to change them causes total chaos in the LTSpice simulator. I'm not sure why 47uF and 4ohms is the right combo, but it seems to be working somewhat better than the last design. Can anyone help me make sense of this a bit?

[Xena E]

 What is the context here: what is your goal?

The last schematic shows a bridge rectifier with parallel CR load... a very low impedance load... ignoring that, the output will be much lower than the input, as the signal has to overcome the two diode drops on each side of the bridge.

The input impeedance of the incoming signal is not specified, and the input coupling of the circuit has to be assumed.

If, as it seems you want, to have some kind of signal level amplitude demodulation perhaps a precision rectifier type circuit would be more suitable?

Sorry to be harsh, but if you want help, some relevant detail will have to be furnished.

Regards.
X

[Amish_Fighter_Pilot]

What is the context here: what is your goal?

The last schematic shows a bridge rectifier with parallel CR load... a very low impedance load... ignoring that, the output will be much lower than the input, as the signal has to overcome the two diode drops on each side of the bridge.

The input impeedance of the incoming signal is not specified, and the input coupling of the circuit has to be assumed.

If, as it seems you want, to have some kind of signal level amplitude demodulation perhaps a precision rectifier type circuit would be more suitable?

Sorry to be harsh, but if you want help, some relevant detail will have to be furnished.

Regards.
X

You're not being harsh at all. In fact, you've helped solve my biggest problem! I swapped out the generic diode models for Schottky diodes instead. I changed a few other values, but that immensely improved my output gain! Now I have to figure out impedance matching. I am new enough that I am not always sure how to ask things. I'm just trying to make a circuit that will render an audio tone from mixing two input sources. I basically want a full wave rectifier with a low pass filter on it, but I am trying to get around using any inductors if possible. Currently the two signals go in and I get two output frequencies: the difference, and half the difference subtracted from the highest frequency. No clue how to show that mathematically yet. I'm learning as I go here.

[Xena E]

Ok. Well if youre presently happy with the topology then one improvement I would suggest is to increase the load resistor by a couple of factors, and reduce the capacitor similarly.

It may be worth your while investigating "active rectifiers".

X.

[tggzzz]

Envelope detection is a term that can have many different meanings. We don't know which one you mean.

The first thing you should do is draw the input or inputs you expect and the output you expect. That includes voltages and frequencies, and the type of modulation.

Depending on those, it may or may not be relevant to understand the difference between "large signal" and "small signal" behaviour. "Large" is, say, 600mV and higher. "Small" is <<600mV. Some forms of envelope detection are based on large signals, other forms are based on small signals superimposed on a DC bias.

Finally, it makes it easy for other people (and possibly you) to understand what you are trying to do with your circuit if you draw it in standard ways. For example, if you are building a four-diode full rectifier, then draw the diodes in the traditional square pattern. Another example is that, in your first circuit, gnd at the top implies that all other voltages are negative. Finally, plonking components down anywhere and then connecting them up is about as helpful as someone writing an essay by plonking paragraphs (or words) down anywhere on a page, and then connecting them with arrows. Both work, but are hard work for a reader.

I did have to do the latter, but that was when "cut and paste" operations were literally cutting paragraphs and gluing down them in the right order on backing paper. Things have improved since then