How to simulate the "on" and "off" of diode in microwave frequency?

[maxsidou]

hello all, recently i'm simulating the diode ring mixer. the "on" and "off" state is often used to analyze the mixer. I'm curious how to judge the on and off state of the diode, then i simulate them in ads, and compare the power of the ends of the diode. But something strange happened. usually, the diode is turned on with saturated power. what i see through simulation is that with the increasing input power, i.e. the diode is gradually saturated, the diode shown more like "off"…… in the following figure, the difference between the ends of the diode increased. IF the diode is "on", should't it be decreased? having considering for a long time but in vain, any idea will be appreciated.

 

[BigBoss]

What is the purpose of this set-up ?? To simulate the diode behavior ?
Since the bias is not applied, in order to turn-on the diode, you should apply remarkable power/voltage.
But it doesn't make sense to do this analysis.I'd use diode mixer configuration then analyze the whole system.

 

[maxsidou]

What is the purpose of this set-up ?? To simulate the diode behavior ?
Since the bias is not applied, in order to turn-on the diode, you should apply remarkable power/voltage.
But it doesn't make sense to do this analysis.I'd use diode mixer configuration then analyze the whole system.

hi, BigBoss, very happy to see your reply. I'm simulating diode ring mixer and its schematic is shown below and i want to design a matching network between LO balun and diode. so i use the above set-up to simulate the input impedance of diode. you mentioned remarkable power/voltage should be applied to the diode. the ring mixer design is passive and the saturated LO power(very high, for example, 13dBm) make the diode "on", by the way, if directly analyze the whole mixer system, how do you judge the matching between diode and balun. in my thought, balun characteristic is affected by its loads. looking forward to your reply, thanks vey much~

 

[vfone]

In the places you mention, only ideal matching networks will work. Any real life reactive components will add a different phase shift in the circuit (due to inherent tolerances). This phase shift mismatch will affect the symmetry and the performances of the balanced mixer.

If look for lower LO level in a (Schottky) diode mixer, a DC bias applied to the diodes can decrease the necessary LO level to about 5dB:

http://www.hp.woodshot.com/hprfhelp/4_downld/lit/diodelit/an995.pdf

 

[BigBoss]

First, let's clarify this point : LO driven Diode Mixers work as in switch-mode so small signal is not a concern here.
Second point, since the small signal working case is not a concern here, a hard driven diode will exhibit very variable impedance depending on signal amplitude.So, if an impedance matching is desired, analysis will definitely be Large Signal mode, not small signal.
In fact, a matching circuit essentially is NOT practically useful for Diode Mixers as well as Gilbert Cell or other kinds.Because is the signal amplitude is sufficiently high, matching circuit will not add an extra benefit for the mixer.Besides, Gain vs. LO Amplitude function/response is almost flat beyond a certain LO driving amplitude point.
Briefly, a Matching Circuit is not essential for LO Port.

 

[BradtheRad]

A diode has some portion of resistance in its PN junction, and some in ordinary ohmic resistance. As you increase current, the PN resistance goes down, and ohmic resistance becomes prominent. At an extreme level ohmic resistance dominates, and the V-versus-A response is almost linear.