smile

You are welcome: https://www.passiveplus.com/hiqproducts.php

https://www.macom.com/products/diodes/schottky-mixer-and-detector-diod ?

https://www.macom.com/products/product-detail/MA4E1310

Than you know that you overpaid for your VNA :)

What??? Buy this one http://www.ommic.com/produits/w2183g-65 or ask them, they will design after your specs.

Make sure your diodes can handle this voltage.

+ amplifier

:bb-offtopic: Is it me, or your amplifier is potentially unstable around 3GHz?

Add resistor(s) in series (or in parallel) to output (input, or both) of the transistor and I am sure you'll be able to match. Everything is possible. It is always a question of a trade-off. Theoretically speaking, simultaneous power matching at the input and at the output exists ONLY IF THE TRANSISTOR IS STABLE. Here is the book about this, read Chapter 3: rftoolbox.dtu.dk

This one is quite broadband:http://ieeexplore.ieee.org/document/419866...98665&tag=1

1. Click on the graph (just to make it active). 2. Press Home/Copy (or just press Ctrl+C) 3. Paste in the text file. All the data will be there.

It's quite broadband, but the noise figure is high and and linearity is poor. This indicates that the phase shifter is based on active devices (transistors). It should be possible to make it on discrete components. Theoretically - yes, but there is no information on switching speed in the datasheet. That would be the limit of your data rate.

I would agree with soldat_shveyk, digital systems you would characterize by BER (bit error rate) instead of noise figure.

Indeed, for circular waveguides you can get flanges without a single screw. Look at this transition for example: Can tell you from personal experience, they work perfectly, at least up to 300 GHz.

No. I just don't want putin's troops in my country to "protect" me because I am russian, like it is in Crimea. :bb-offtopic: