Diplexer

Not an overly complicated task but I built this little diplexer yesterday as I intend to get more into LEO (Low Earth Orbit Satellite) activities and plan to build something like an arrow antenna. Originally I planned to purchase one on eBay but was surprised that people pay more for them in an auction than brand new in the shop! How funny is that?

How it works

The diplexer is basically only a high and a low pass butterworth filter with a 3db-frequency chosen at 250MHz. The challenge is to build it as steep as possible so the unwanted frequency is attenuated as much as possible but still keep the insertion loss in a region that does not attenuate the signal significantly.

There is plenty of material available on the net, I found this here pretty useful:

Diplexer from Tonne Software (a very  efficient specialized design tool which allows to play with the values and see the effect)

RFSIM (a general purpose RF simulation tool. Not ideal for a diplexer as it only allows simulation of two ports but still good to see a single filter)

DF6VB's instructions for building a diplexer (which are very close to how mine works - only in German though)

Validating that it works

Sadly my trusted miniVNA only works up to 180MHz so I could only see the low pass up to that point. To validate if the 70cm branch works (which is more critical to calibrate) I just connected the common port to the frequency generator and the output to the wattmeter and adjusted for maximum output.

When feeding a 0dbm signal, the attenuation in db can directly be read on the wattmeter display. When feeding a 2m signal (146MHz), the insertion loss on the 2m port is 0.1 to 0.2 db, the attenuation on the 70cm port is -55db.

When feeding a 70cm signal (440 MHz), the insertion loss on the 70cm port is 0.1 to 0.2 db and the attenuation  on the 2m port is -46 db. Good enough for me!