Some years ago, thanks to some loaned equipment from my friends Bob G8DTF and Dave G4MVU from the Bolton Wireless Club I made some contacts on the 9cm band.
Initially I was using Bob's downconverter (with a 144MHz IF) and his varactor-multiplier transmitter with my own SNA586 MMIC as a "PA" - but the x8 frequency multiplication (from 425MHz) meant that we were limited to "not-so-narrowand FM".
Then Dave loaned me one of his linear transverter systems, in which he'd converted a couple of AirSpan AS4000 WLL units to use a 23cm IF, which enabled us to make clear SSB contacts from Winter Hill down to his QTH.
I was so impressed with this that I built a similar system for myself. But I used an IF of 880MHz (very close to the 850MHz originally used in the AS4000), which saved me having to change the VCO inside the unit - although I had to make a crude 880MHz "transverter" to go along with my 2m transceiver.
Here are some pictures of the project in progress:
I field-tested this during the August 2012 SHF UKAC session: my first QSO was appropriately enough with Dave G4MVU and signals were S9+20dB both ways.
Then after using it portable over the summer of 2013, I loaned it to Mark M0UFC to use in the SHF UKACs; he used it with his FT817 rather than my IC202.
In 2017 I bought one of the new C-band LNBs with a PLL oscillator. This downconverts 3400MHz to a "negative IF" (i.e. frequency-reversed) around 1750MHz, which can be tuned with a FunCube Dongle or some of the RTL-SDR units (ones with the E4000 tuner).
With this bolted onto my mast, despite an obstructed view to the South I was still able to detect a beacon (probably GB3ZME from Shropshire) around 3400.9MHz. My waterfall display shows the two frequencies of the beacon's FSK signalling, along with a slow HF frequency drift caused by my LO heating up and a more rapid LF shift due to the Doppler-shift from an aircraft reflection.
I had set the "Swap I & Q" option in HDSDR to compensate for the "negative IF": therefore frequency increases from left to right on my display, and I can receive USB signals on "USB".
My original tests used the supplied "flat scalar" flange around the LNB, which is designed to provide a wide beamwidth to illuminate a prime-focus dish. Then I built a crude conical horn to give better directivity (folded from 0.5mm sheet aluminium and secured in place by a pipe-clip). I've recently compared the radiation patterns between the two.
I tried it fastened higher up on my mast too and the beacon was clearly visible all the time, in brighter colours on my waterfall display - and I could even hear it through the laptop speaker!
In 2022 I revisited this system, this time using a HackRF One to tune the IF. However found that I could no longer detect the beacon, but instead saw lots of interference on the waterfall and a local CW signal source showed horrible sidebands (at ~200 Hz and harmonics). Initially I thought that this was a parasitic oscillation within the LNB, but then discovered that it only happened when the LNB was up on the mast - and the effect got worse when I pointed it South! I soon deduced that it was stong local signals in the new "5G" cellular band (recently re-allocated from the former Amateur band) which were overloading the LNB and producing all sorts of cross-modulation.
At first I thought that this spelled the end of the LNB's usefuless, but then on inspecting its innards I saw that it would be easy to bypass a MMIC amplifier and reduce its pre-mixer gain. Having done this, all the cross-mod disappeared and I was able to see clean signals on the band once more.
I also used it to make this wideband spectrum plot (in max-hold mode) to show the local cellular signals in the "n78" band between 3410 - 3800 MHz.
No doubt I've reduced its sensitivity, but I'm still able to detect a distant beacon (probably GB3OHM hear Birmingham) by rain-scatter and monitor the local stations during the Tuesday night SHF UKAC events. And in June 2022 I even heard G8DMU/P (in IO94) coming in on a back-scattered reflection from the South!
The reverse-IF makes it difficult to work out the actual frequency to which the receiver is tuned, so I've developed a Python Tkinter GUI to help. First I need to tune the HackRF One to 5.15 GHz and measure the LNB's LO frequency directly, then come down to 1.75 GHz and subtract the IF. But the SDR's frequency-reversal also needs to be accounted for - this effectively "mirrors" the spectrum.
After the receiver section of my original AS4000 transverter died, in 2020 I discovered that I could use my ADALM Pluto SDR unit to receive the local stations - just connecting my dual biquad antenna straight into its Rx port, without any LNA! For a while I used a hybrid system, running the SDR Console program and feeding the Pluto's Tx at 880 MHz into the transverter's IF - but this was very tricky to use, as it took a while to manually net my transmit frequency onto each incoming signal (before they QSY'd).
Then I bought a small driver amp to take the Pluto's output up to ~300mW and this replaced all the AS4000's innards, feeding into the original 8-element patch antenna array (which I estimate has a gain of around 15dBi). This has removed the 880 MHz IF, so now I can automatically net directly onto any incoming signal.
This system is compact and lightweight enough to be deployed from an upstairs window, to reach local stations to the East or West. And the latest version will mount half-way up my main mast to aim to the South too (and is a little more weather-proof, thanks to a plastic food container for the Pluto and a "poly-bag radome" for the transmit antenna!)
