how to design a C-band Noise Source?

How to design a C-band noise source?

Can anybody give a suggestion? Attached is the picture. Could anyone give more explanation about the diode and trace length?

**broken link removed** Pls download SECTION A DIODES Catalogue **broken link removed** and look pp 21-38. I think that their NS-303 are NC303 in BL Package **broken link removed** from **broken link removed**

This is only up to 2.5GHz, but gives you an idea how to make it.

Designing a noise source is not an easy task.
One has to find a suitable noise diode for the frequency band, measure its impedance under specified conditions (usually not given by manufacturers) and design a matching circuit for an expected performance.
To test the design, one needs to use a good calibrated spectrum analyzer or a sensitive receiver; another good idea is to have a calibrated noise source for comparison.
For those who want to try an interesting work: OK1TAY invented that the -e-b diode of a RF SMD transistor (can be extracted from an old PCB for free) can be used in avalanche breakdown, and can generate ENR~30 dB up to 12 GHz.
I have tried it and operated many such low-cost noise sources, up to 12 GHz.
Use a DVM and connect the e-b diode to a 12-15 VDC source with 1-2 kOhm resistor in series. A suitable diode breaks at 6...8 V, some at >10V. Then add a resistor attenuator, ~10 dB, and connect the device to a spectrum analyzer. You will see that DC current shapes the spectrum. Adjust it as needed, then use the noise source.