Frequency doubling using diodes

[Aiy]

Hello Everyone,
As some may have seen I have a broken Anritsu SDM module (Topic here: https://www.eevblog.com/forum/repair/wiltron-anritsu-switched-frequency-doubler-module/ )
This module has a frequency doubler which mutiplies the fundamental 2-20GHz YIG oscillator in the Anritsu machine to give output up to 40GHz.

My question is this:
If I attempt a repair and switch out the diodes, can I in principle use any Schottky diodes with enough power tolerance to repair my unit, or is there something else to it?

I see that Macom has the step-recovery type listed for multipliers and comb generation (possible candidate would be https://www.macom.com/products/product-detail/MMDB45-B11 ), but my knowledge is lacking in this field.

The original diodes in the SDM device seem to be four diodes in series, anchored in between the middle. The whole thing is a beamlead thing that I have not found anywhere, so its probably custom..

ASCII CAD: ( ->| = diode )

GND - 50R - ->| ->| - Sig ->| ->| - 50R - GND

I have measured the forward voltage drop between the "inside of the 50Ohm resistors to the "Sig" point, and it's 1.2V-isch, indicating two diodes.

Pictures of the SDM unit is in my other topic..

Best regards

[Joel_Dunsmore]

posted in the link instead of here:
If you happen to attend the ARFTG / RWW conference starting next week, you'll likely find Jon Martens there, and I guess he might have either designed the part or knows a lot about it.

www.arftg.org or www.radiowirelessweek.org

tiny diodes, likely with very low parasitic capacitance.

[RoV]

I think they are GaAs Schottky diodes, a couple in series each side to handle the required peak voltage.
Referring to the other post of yours: the input 20 GHz signal enters from the top in microstrip; then there is a transition from microstrip to slotline: the diodes rectify the signal and send the opposite half waves on opposite sides, so the signal that propagates along the slotline is a full-wave rectified version of the original, hence at double frequency. Follow matching and filtering stages, probably based on varactor diodes with bias adjusted according to the frequency. The end of the slotline seems to be just below the main line of the 1st figure.
Replacing those diodes is really hard...