Michael Margaras SV1CAL – Amateur Radio and Electronics

dish_july2020

Parabolic reflector for weather satellite reception on 1.7 GHz and Moonbounce on 1.3 GHz

Note: Need to make a nicer, balanced counterweight. That ugly full of concrete box was a quick workaround to create some balance 🙂

The 23cm setup consists of the following :

For HRPT weather sat reception , see here

1) 2.5 m mesh dish , f/D=0.4

2) Big RAS SPID rotator

3) Controller is SPID’s Rot2Prog or homemade “EME FUN antenna controller”, see related article

4) Tracking is done by “EME FUN software”, see related article

5) A dual dual mode septum feed is used based on the W2IMU / N2UO design

6) Home-made power amplifier

6) G4DDK LNA kit applying all RW3BP’s suggestions : SMA male connector, removed DC blocking cap . NF measured <0.2 dB ( PXA analyzer and a N4000A noise source)

7) Home-made 144/1296 MHz up/down converter

How to improve sensitivity by more than 1 dB by improving isolation between TX and RX ports:

The plot below shows the Measured Isolation between TX and RX ports with a E5071A network analyzer.

tx_rx_isolation

The feed alone looking at cold sky (red trace) results in better than 45 dB TX/RX isolation , but when the feed is in the focus of the dish (in my case f=1m) the isolation becomes only 13 dB (blue trace) . From theory we know that Noise power is proportional to the temperature of an object . Assuming a perfect match and 27C which is 300 K at the TX port, 300 divided by the isolation , thus noise power which corresponds to 15 K will reach the LNA port (RX) .

For a W2IMU feed and a 0.4 f/D dish minimum antenna noise temperature is close to 10K, a 0.2 dB LNA adds 14K and the coaxial protection switch 3.5 K ( assuming 0.05 dB loss ) thus totally 10 + 14 + 3.5 =27.5 K .

To this noise temperature one has to add the 15 K coming from the TX port.

So the total noise temperature is then 27.5 + 15 K =42.5 K.

Implementing RW3BP’s excellent suggestion, to improve TX/RX isolation a metal disk of about 73mm diameter has been placed in the dual mode section of the feed.

The exact position in the dual mode section and the diameter had to be found by trial and error. It is relatively easy though to trim dimensions in order to achieve 30 dB of isolation, practically eliminating the noise coming from the TX port.

For the numbers above, the sensitivity benefit will simply be 10*log10(42.5/27.5 ) = 1.9 dB

The lower the system noise is (dominated by Tlna+Tant) , the larger the benefit will be .

In my setup, at least 1.2 dB higher sun noise/cold sky could be achieved by just placing that small metal disk.

Running an optimization of the entire setup ( dish , feed , coin ) of the coin/metal disk dimension and position, revealed that the values found by trial and error were very close to the simulated ones.

On larger dishes where the feed is further away from the reflective surface , isolation will also be greater and as a result less noise power will be coupled into the RX port.

Plot below shows simulated isolation for a 6 m dish vs f/D . For 0.4 it’s about 20 dB.

6mDish23cm_TXRXiso