Having posted my setup at the end of last month and seeing that my mail program wrapped line feeds, decreasing readibility, here again my description and perspective.

Es'hail2 will not differ from any other geostationary broadcast satellite in a sense that it's a 'full duplex' repeater with some time delay.

I reckon this time delay inhibits actual full duplex operation, just like the TV reporter for a 'live stand up' in the TV news via a satellite link having an earplug with all audio except himself ('N-1').

For us amateurs this means no (significant) Doppler shift, no LSB/USB issues etc.

Just point your dish once to 25.5E like (more than) thousands of broadcast satellite viewers do (without technical knowledge at all ;-)

Downlink (10.5 GHz): For narrow band usage frequency and stability is paramount. For 10.5 GHz we're helped by the BC satellite industry, offering cheap PLL LNB's with 'HDTV precision'.

Imho this precision needs to be increased by injection locking the LNB PLL with e.g. a GPSDO due to the large multiplification factor inside the LNB (e.g. for Octagon 361.111111).

Noise figures (NF) of these LNB's are somewhat higher on 10.5 GHz due to (steep) wave guide cut off, but still usable, compared to the ease and price.

Uplink (2.4 GHz): A 10W BUC (block up converter) suffices with published specs of the 2.4 GHz receiver above in conjunction with a ca. 75cm dish here.

I'm currenly developing a simple 2.4 GHz BUC and will use old UMTS pallets to amplify the desired 2400 MHz output. Picture of my UMTS pallets: http://remco.org/umts.jpg

Bear in mind, S-band power is cheap :-)

Antenna: One dish, the larger the better. Not for the uplink, but for improving dynamic range at the receiver (downlink) side.

I will use a surplus Triax 88cm dish (for portable / 'umbrella' use ;-), f/D is around 0.6 (like most consumer BC satellite dishes).

Feed: From my perspective the only challenge lies in the 2.4 GHz uplink. That is, to make an S-band feed which phase center coincides with the 10.5 GHz LNB (which has its own 'feed' ;-)

I've no fear for isolation issues due to the large frequency difference.

I am told that circular polarisation (CP) is 'mandatory' for S-band uplinks in general but can't find a verification for this, perhaps somebody else?

Anyway, we're amateurs and creative. So when RHCP uplink is too difficult we make a linear antenna and increase uplink power with 3 dB ;-)

But, I follow the published receiver specifications . . . Last week I made a 6 turn LHCP 2.4 GHz helix. (Rule of thumb seems to be 1 turn per 0.1 f/D) Phase center of such a helix is a bit 'foggy' but from what I read lies somewhere between the first and second turn.

I've some issues with matching it due to some mechanical boundary conditions of the helix feed point in conjunction with the mounted downlink LNB and relative position of the helix. There is an additional 'transmission line' involved and a 'normal' 1/4-wave stub doesn't give me the desired return loss. So, this needs some fiddling, perhaps with a gamma match or so.

My approach is, the downlink LNB is mounted in the dish 'as usual' and looks through the centre of the uplink helix.

Pictures of the prototype feed arrangement can be provided soon. They are on a camera not present here at this moment of writing.

Remco PA3FYM