How to measure the S-parameter of schottky diode?

[samtsui]

schottky

Hi
I would like to ask how to measure the S-parameter of schottky diode?
is it just connect the anode pin at to signal pin and the cathode pin to ground at the network analyzer?
Sam

 

[biff44]

Re: RF schottky

You can consider a Schottky diode to be a large signal device. That is, its impedance can vary drastically as the RF power incident on it increases.

If very little power is incident, the diode is reverse biased and looks like a very small capacitance. If enough power is incident, the diode conducts current for part of the input sine wave, and the diode now looks more like a larger capacitance in parallel with a resistance.

I would set up a bias tee structure (possibly inside of the network analyzer itself) so the DC component has a resistive load on it. I would then pump as much RF power into the diode as possible and make a measurement. Remember you need the incident power expected during operation, plus more power added to conquer the mismatch loss (assuming you are going to match the diode later on).

 

[samtsui]

Re: RF schottky

Thanks biff44,
so if i see the real part impedance is very small (about 1-2ohm), is that mean the power is not enough?

 

[biff44]

Re: RF schottky

You probably have a reflection coefficient that is almost 1 in magnitude. You can think of that as a small resistance (say 2 ohms) in series with a small capacitance (say 0.2 pF), or as a large resistance (say 2000 ohms) in parallel with a small capacitance. But, since the reflection coeeficient is almost unity, you can not build a matching network for that.

But if you drive the diode with enough RF power so that it starts rectifying, then you get a better guess at the diode's impedance to match to. This is the "large signal" impedance you want to design the matching network for.

By the way, one common "matching" technique if you do not care about efficiency, is to place a DC blocked 50 ohm resistor in parallel with the schottky diode. You have a broadband low reflection coefficient detector now--albeit not a very efficient one.

 

[samtsui]

Re: RF schottky

If i bias the diode with a dc voltage through a inductor, and connect the network analyzer with a dc blocking cap, for this setup can it measure the S-para correctly if the network analyzer cannot provide a large power level during measurement?

Thank a lot