input impedance of antiparallel diode pair

[elektr0]

Hello,

I need a second opinion.
So, simple question.
Assume a diode, which has 50 Ohm input impedance at a certain drive level / frequency / embedding network.


If we take two of these diodes in parallel. We should change the input circuitry to 100 Ohm and double the input power.
If we take two of these diodes in antiparallel config. What input impedance do we have to take (again: twice the input power) ?

Thanks
-e

 

[jiripolivka]

Diode RF impedance is function of input power. Antiparallel Schotky diodes pumped by 10 mW approx. do have the input impedance close to 50 Ohms.

 

[elektr0]

Hello jiripolivka,

the diode impedance depends on pump level and the diode parameters (barrier height). There is no general drive level for 50 Ohm impedance.
But the question was, do antiparallel diodes have half the large signal impedance than a single diode.

-e

 

[jiripolivka]

Hello jiripolivka,

the diode impedance depends on pump level and the diode parameters (barrier height). There is no general drive level for 50 Ohm impedance.
But the question was, do antiparallel diodes have half the large signal impedance than a single diode.

-e

I never tested the impedance of a single diode. In a balanced mixer I confirmed that the two diodes in the RF cavity are connected in series and exhibit close to 200 Ohms; in the LO circuit they are in parallel and exhibit ~50 Ohms under 10 mW pump power. In a diode tripler an anti-parallel connected pair has ~50 Ohms but perform best under ~50 mW pump.

All real experience points to that RF power strongly affects diode impedance but in different circuits the real impedance differs widely. So the question cannot be answered technically straight.

Diodes in detectors (single) are generally mismatched over all input power range. Improvement is possible by DC biasing (replacing pump power).

 

[elektr0]

Hello jiripolivka,

imagine a single diode, which has large signal input impedance Zd at a single frequency and drive level P.
Putting two of these diodes in parallel, leads to Zd/2 at 2P.
What would you expect if two of these diodes are connected in antiparallel configuration ?

-e

 

[jiripolivka]

Hello jiripolivka,

imagine a single diode, which has large signal input impedance Zd at a single frequency and drive level P.
Putting two of these diodes in parallel, leads to Zd/2 at 2P.
What would you expect if two of these diodes are connected in antiparallel configuration ?

-e

Your idea may work at RF. My experience with mixers and multipliers above 10 GHz indicates that reality is far from models. You can try a real circuit and measure it.