2007-07-14
Non-Emissive Airband Receiver
I was looking into building a non-emissive receiver for the air-band to take with me on the odd occasion I take air transport. Obviously with all the EMC paranoia with aerospace technology the only "safe" device is one with no possibility of RF emission, even at microwatt levels so regenerative and super-regenerative receivers are out.
Charles Wenzel's excellent site has an interesting circuit along these lines. Essentially an amplified VHF crystal set. It uses a biased 1N5711 to achieve good sensitivity (rather like my wavemeter), and a front-end resonator to make it sufficiently selective in the band of interest (instead of being wide-open like his Amazing All-Band Receiver).
The circuit is quite trivial, it took me only minutes to build on a solderless breadboard. I used a TL072 opamp rather than an LM358 (Which makes very little difference at all, despite having many orders of magnitude better input bias current and about three times the unity gain bandwidth. Its main advantage is its rather low noise.). Once built it works very well indeed. The selectivity is limited almost completely by the Q of the inductor in the front-end, and is acceptably wide for the application. I tried tapping down on the inductor (a few turns of tinned copper wire) to affect a better match into the detector, the improvement in selectivity isn't really worth the reduction of sensitivity.
The audio gain is more than sufficient to hear the noise from the resistors in the front-end, it might be worth using less noisy devices actually. The total audio power might be a bit weak inside an aircraft, but for around town it is plenty loud.
On the FM broadcast band I could separate about three of the major high-power-end stations, specifically 2DAY, 2MMM, and 2MIX. 2WS is for some reason the easiest to tune in (at modest distortion), it may be that the circuit Q is near-ideal at 101.7 MHz, or perhaps it is just far enough away from other powerful stations that it is easier to slope-detect without confounding garbage from adjacent stations. 2DAY and 2MMM are fairly hard to separate, you have to tune below 2DAY and above 2MMM to avoid the other. 2JJJ for some reason isn't sufficiently powerful in my shack to work well with the receiver, but you can understand the speech of the announcer. 2ABC is also easily detectable, but the quality of the demodulation is not fit for its programming.
You certainly would not want to use this receiver to listen to FM broadcast, it sounds even worse than a super-regenerative receiver (and is hopelessly deaf), but perhaps with some effort in constructing a helical resonator it might be improved to the point where it was acceptable. I still think the best non-FM receiver I've built for FM broadcast is the regenerative one, which has the advantage of variable-Q and good sensitivity.
On the Airband I couldn't hear a thing with the receiver. It isn't sufficiently sensitive to hear off-air signals, but experiments were carried out with the signal generator suggests it should work fine in close proximity to an airband transmitter. It works quite well on 2 metres to hear my fredbox, and can separate it from the pagers above 2 metres. It can hear TV stations very well, especially ABN-2 and ATN-7. With a baseband amplifier of sufficient bandwidth I'd imagine it would be possible to actually watch TV (poorly) with the recovered signal!
Thoughts
Such a circuit is extremely useful around the bench. Much as Charles has built a general purpose one as a detector I am considering building one with pluggable coils to cover MF-VHF. It would not be extremely difficult to add the audio amplifier stage to my existing wave meter (currently it just has a Hi-Z headphone output).
For the original purpose of air-band reception at an airport or onboard an aircraft I think a better solution might be a TRF receiver. Well, a TRF receiver with a bandwidth of about 20 MHz! Reviewing the noise-floor math, suggests it would be quite practical to simply amplify the entire airband (after a good front-end filter) and AM detect that (maybe with an AGC loop). The thermal noise floor is around -98 dBm, which means it will need signals significantly more than 3 uV for a reasonable SNR assuming a perfect noise figure. At 20 uV the signal will need around 80 dB of power gain for a mW of audio (around 200 mV into a 32 ohm mylar speaker - quite loud). With 40 dB each of RF and AF gain this seems quite do-able with stability. Making the receiver in this manner means it requires no tuning. (i.e. you hear *everything* in band, at once). The big problem would be VHF beacons dominating the reception, but perhaps a tunable trap or two in the front-end would let you notch out any that were causing you trouble. Beacons are allocated in the 1st 10 MHz of the band, which could be rejected by the front-end BPF.
The BPF for such a receiver would need a loaded Q of only 6 at a harmonic centre of 127 MHz, so it is quite practical to build. Two resonators loosely coupled should do the trick, 100 nH and 15 pF is about right with an impedance tap about 1/3 of the way up from the cold end. 22 pF trimmers could be used to tune, or some squash/spread of the 6 turn coils (8 mm ID, 15-20 mm long). Anyway, that is back of the envelope stuff, only way to tell for sure is to build it. The RF amplifier is the most difficult part, 20 MHz wide at VHF, low noise, stable and preferably AGC-able with around 40 dB power gain.
2 comments.