AM demodulator (envelope detector) diode selection

Thread Starter

Catriona82

Joined Feb 15, 2017
42
Hi guys,

I'm looking to make a very bobby basic envelope detector out of a diode and RC network. I modelled two circuits with different schotty diodes and got very different waveforms (see attached). One matched the input envelope well but had some high frequency noise on the output. the other was smoother but had a very slow decay on the falling edge.

I would rather have the former but I don't know what diode characteristics to be aware of when selecting. I thought it might have been the capacitance when reversed biased but for two of models I used that had slow decays, one had a larger capacitance and one had a lower capacitance than the MBRS340 that I used in the "good" model.

I want to have as low a forward votlage as possible as my signal is quite small.

Does anyone have any insights they can share with me?
 

Attachments

Thread Starter

Catriona82

Joined Feb 15, 2017
42
Why is that? I'd like to get some understanding to the reasons for choosing the characteristics of the diode rather than blindly picking one.

Thanks.
 

DickCappels

Joined Aug 21, 2008
6,661
Or, you can used a biased Schottky diode for better linearity and better sensitivity.

The cat's whiskers used in the earliest crystal sets were Schottky diodes with an approximately 0.7V forward voltage. Back in the early 1900's they were sometimes used with an adjustable bias voltage source to improve sensitivity.

In the past year I compared some 1N60 germanium diode (also widely used in crystal radios) and compared them with the newer 1N60 Schottky diode and could only tell them apart by their temperature coefficient of voltage for an given input current.

1598473882138.png

Above is a baised diode detector that I used for a broadband RF signal indicator. For a TRF crystal set L1 and L2 would be replaced by a parallel LC tuned circuit.

D2 is only there to provide temperature compensation for the voltage drop across D1. For audio you will AC couple the output and D2 can be replaced with a jumper.
 
The germanium diode doesn't look great as the spec sheet quotes 1V max forward drop at 5mA. A small signal Schottky diode is the way to go as this offers lower forward drop and hence better sensitivity. The BAS70-04 suggested above is a good candidate with 410mV max forward drop at 1mA, as is the RB705D you chose. Going by the curves on the datasheet you get 600mV at 5mA for comparison with the 1N32A.
The reverse capacitance effects depend on the carrier frequency being used. The difference in the two circuits you show is more to do with the reverse leakage of the high power Schottky diode. The way to fix the droop is to reduce the value of R1 rather than rely on high leakage current from a power rectifier.
Here is a wide range RF level detector I have used on several protects that attempts to fix the non linearity of the diode detector. It does give good input signal voltage range.
Diode detector.png
This circuit needs an opamp that operates with inputs at the negative rail (pnp or p channel inputs) but these are common enough.

Chris.
 
Top