Dear colleagues
as a fun and small christmas project, I built a little noise source for my 8970B NF meter.
I borrowed a HP 346C noise source which is known good and used that to calibrate my setup.
Then, I made this little prototype in a small aluminium box. I used the schematic published here
http://www.simonsdialogs.com/wp-content/uploads/2015/08/noise-source-bfr93a-shunt.jpgbut I used a BGF403W transistor.
As one can see from the plot attached, the ENR is a bit high, even though I have used a 10dB pad inside. However, at higher frequencies, it seems that the ENR is quite flat. But I guess from adjusting the output coupling capacitor, I can somehow optimise the frequency flatness. From the fact that for >1 GHz the flatness is quite good, I assume that this noise source would work well for even higher frequencies. I do have an 8971C noise figure test set which works up to 26.5 GHz, but I have not measured my source with that yet.
Later I plan to make this noise source with a nicer case. Compared to the noise source electronics, I am even more proud of my prototype case I milled these days. For comparison, I made a photograph otgether with a HP 8484A power sensor such that the size and shape of my own case can be seen. I plan to make a bunch of these cases and also use them for my own built power sensor heads (compatible to HP 436A).
To make the actual noise circuit a bit more reproducible, I plan to make a small PCB which holds the transistor and stuff. As one can see from the photographs attached, I use these types of N connectors with a flange, and I wonder how these connectors are meant to be connected to something?
On the opposite side of the N connector, my case has some space to mount the connector required for the 28V noise source power supply.
Home
Help
Search
Login
Register
