Simulating a simple envelope detector circuit

[tsunami_all]

Hi,

I am trying to simulate a simple envelope detector circuit on TINA simulation software.
Here is the schematic:



I supplied sine wave of 2.45 GHz and 1V amplitude. I am not getting the output waveform as expected.



Any idea what might be the issue here?

Thanks

 

[barry]

What do you think the issue is? It looks right to me.

1) Why are you using THAT Schottky diode? It's got a lot of reverse leakage voltage.


You need a different diode. Try that circuit at a lower frequency and you'll see it works.

 

[tsunami_all]

Hi,

Which diode part number do you suggest?

Thanks

 

[KlausST]

Hi,

"

I am not getting the output waveform as expected.

Then you need to tell us what you expected..

Klaus

 

[tsunami_all]

Hi,
This is the expected output waveform.
I am expecting output waveform with its negative side removed. The waveform should charge and discharge according to the RC values.

Thanks

 

[betwixt]

Feed it from a line transformer at 50/60Hz and that's exactly what you will get.
Feed it from 2.45GHz and the power Schottky diode will just look like a capacitor. Use a small signal diode or better still, one designed for microwave frequencies.

Brian.

 

[danadakk]

The 1.5M Ohm R is not going to draw much charge off 1 nF cap,
so you would see, if you take sim out further in time, more of a
DC result with very little ripple I would think.

The RC time constant is ~ 1.5 mS, so it that time you have had a ton
of 2.4 Ghz cycles to keep C charged, once its initial ramp up complete.

Also, I am not a Tina expert by any stretch of the imagination, but
doing trying to do a 2 Ghz sim with it seems very aggressive. Especially
in light of Z matching, trace effects (transmission lines), etc.. ADS from
Agilent is what I would think of to do this type of sim.

Regards, Dana.

 

[d123]

Hi,

I have never used an envelope detector or even simulated one before today. It's been an exciting day for me. I do use Tina regularly for simulation purposes, 'though. It's a shame you neither show the timescale nor the VG2 internal resistance you chose. I simulated your exact circuit with VG2 as having an unrealistic 0 Ohms internal resistance and by trial and error I found that changing the 1nF capacitor to 10pF makes the output follow the input very faithfully with a small loss in amplitude. In that magical place we all love called Simulationland.

You may want to do a piecewise linear of your own as well as a sinewave to see how well it follows the input signal, unless you only expect to see a sinewave like AM modulation stuff.

I'm not very convinced by post #5 'expected output waveform', in my ignorance, my very great ignorance, that looks like a bog-standard power supply rectified and smoothed signal more than an envelope detector. Why do you think a signal that should faithfully follow the input should hold up the output voltage? I'm hoping to learn something about envelopes, that's why I ask.

And, surely for it to have no negative level at the output requires a second diode that is reverse-biased?

 

[betwixt]

d123, at risk of hijacking the thread, that isn't what an envelope detector does.

The intention is to recover a low frequency amplitude modulation signal from a high frequency carrier signal.
The modulating signal changes the amplitude of the high frequency carrier, for example a 1KHz sine wave modulating a 2.45GHz carrier would make the carrier level rise and fall at a 1KHz rate. An envelope detector recovers the 1KHz when fed with the modulated 2.45GHz signal. The concept of 'envelope' comes from the outline of the carrier shape rather than the shape of individual cycles within it.

The problem with the schematic is just that the diode chosen is not suitable for use at 2.45GHz. The junction capacitance alone is about 50pF which looks to the signal rather like a 1.3 Ohm resistor at those frequencies.

The way it is supposed to work is the diode rectifies the signal and the RC time constant of C1 and R1 recovers the envelope shape. There will always be some residual ripple in the output but careful choice of time constant will let C1 charge and R1 discharge the voltage at a suitable speed that the amplitude of the carrier is closely followed.

The 1N5819 is a power rectifier, a small signal Schottky diode or a dedicated microwave diode would be more suitable. Bear in mind that the frequency is high enough that wiring inductance and stray capacitance become significant factors if it is constructed.

Brian.

 

[d123]

Hi,

Wow, not only did I get the general idea very wrong of what the output needs to look like in my previous post, so my apologies. I must say: 'simple'? From what I'm seeing, envelope detectors are anything but simple...

The best I've simulated so far today is this, at a pathetic 10MHz, and it is far from faithful to the amplitude of the input signal:

That looks very hard to get right, even only in a simulation..., and ignoring the poor choice of components selected.

--- Updated Jan 26, 2021 ---

Hi Brian,

d123, at risk of hijacking the thread, that isn't what an envelope detector does.

The intention is to recover a low frequency amplitude modulation signal from a high frequency carrier signal.
The modulating signal changes the amplitude of the high frequency carrier, for example a 1KHz sine wave modulating a 2.45GHz carrier would make the carrier level rise and fall at a 1KHz rate. An envelope detector recovers the 1KHz when fed with the modulated 2.45GHz signal. The concept of 'envelope' comes from the outline of the carrier shape rather than the shape of individual cycles within it.

The problem with the schematic is just that the diode chosen is not suitable for use at 2.45GHz. The junction capacitance alone is about 50pF which looks to the signal rather like a 1.3 Ohm resistor at those frequencies.

The way it is supposed to work is the diode rectifies the signal and the RC time constant of C1 and R1 recovers the envelope shape. There will always be some residual ripple in the output but careful choice of time constant will let C1 charge and R1 discharge the voltage at a suitable speed that the amplitude of the carrier is closely followed.

The 1N5819 is a power rectifier, a small signal Schottky diode or a dedicated microwave diode would be more suitable. Bear in mind that the frequency is high enough that wiring inductance and stray capacitance become significant factors if it is constructed.

Brian.

Thanks for the explanation. 'Simple', I think it is not as a circuit to make, by the look of it. And cringe big time, in my case.

 

[barry]

I'm not very convinced by post #5 'expected output waveform', in my ignorance, my very great ignorance, that looks like a bog-standard power supply rectified and smoothed signal more than an envelope detector. Why do you think a signal that should faithfully follow the input should hold up the output voltage? I'm hoping to learn something about envelopes, that's why I ask.

And, surely for it to have no negative level at the output requires a second diode that is reverse-biased?

View attachment 167156

The OP is not looking for a rectifier, he is looking for an envelope detector. His 'expected' waveform in post #5 is, in fact, what the output of an envelope detector should look like. It is not supposed to 'faithfully follow the input', it is supposed to charge the cap to the peak amplitude and hold it. You've changed the circuit into something completely different.

Assume an ideal diode, an ideal cap and no shunt resistor. On the first cycle of the waveform the cap will charge to the peak value of the input and stay there for eternity. Adding the shunt resistor will cause the output to decay during the negative-going portion of the input (the diode is reversed biased). The OP's problem is that he doesn't have an ideal diode; he actually has a diode singularly ill-suited for its intended purpose.