Living in the middle of a noisy housing estate can make HF radio reception difficult, or to view it another way, challenging! End-fed wire antennas are not the best for noise rejection when feed point is in the shack. Things can be improved but that means using magnetic baluns, feeder isolators and coaxial cable feed. I decided I’d investigate some alternatives
I’ve looked at loads of designs for LF and MF reception but there seemed to be a shortage for use on HF (3-30MHz). Most of the designs I found were for transmitting applications and I wanted one for receive only. Add to this the debate between wideband loops and tuned loops and things can get quite confusing.
- Narrowband loops – high Q – can generate internal thermal noise – simpler to build – provide additional receiver selectivity.
- Wideband loops – Low Q – higher losses and require high performance matching amplifiers – capable of very good performance.
For an excellent comparison and explanation of the differences see the Wellbrook Communications site. For a narrow-band application such as beacon monitoring then in my opinion the tuned loop wins hands-down.
Thanks to John GM4SLV who showed me a QRP transmitting and receiving loop designed by G4FON. This design also used a better method of feeding the loop – a ferrite toriodal transformer (why didn’t I think of that?). It also used the outer braid of RG58 coaxial cable to provide the loop element, effectively a 4mm dimeter conductor.
To ease the design process and save lots of hard maths I used one of Reg G4FGQ’s excellent programs, RJELOOP2.EXE. This allowed me to adjust the design to suit the components and space available. It also mentioned again the use of a transformer for feeding the loop instead of a coupling loop. Here’s the final design (well, the mathematical model of it!):
This indicates I should get a gain of -14.5dB relative to an “ideal” antenna (dipole or isotropic?). This value is quite acceptable at HF. Comparative loss values for smaller loops:
- 1.0m sides -18.6dB.
- 0.75m sides -22dB.
- 0.5m sides -27dB.
- 0.3m sides -34dB.
Run it through RJELOOP2.EXE and see how they compare.
RG58 coaxial cable for the main loop winding, the outer braid is used as a 4mm diameter conductor. The loop sides are 1.5m long, a total circumference of 6m. This is enclosed in a square loop structure made from 15mm white PVC electrical conduit. The corners are matching conduit corners, circular junction boxes are used at the top and bottom for the tuning capacitor and feedpoint respectively.
The tuning capacitor uses a 60pf ceramic trimmer in parallel with a 100pf silver mica, mounted on a small piece of single-sided copperclad.
The matching transformer uses a FT50-43 toroid. The loop conductor is fed through as a 1-turn secondary and the primary comprises 11 turns of 26swg enamelled copper wire.
So far, quite pleasing. The 14.5dB loss is barely noticeable and the noise level has been reduced from S6 to S1 on my TS-870S meter (preamp off). Brief tests listening on 5290kHz allowed me to hear GB3RAL and GB3WES down to 7 steps. I have now (March 2005) installed this loop as the main antenna on the beacon receiving system.
Some pictures: not very pretty but it does work.
Tuning Capcitors, 60pf ceramic trimmer and 100pf fixed silver mica.
Feedpoint showing toroidal matching transformer
At 1.5m a side the Mk1 loop is ideal for a fixed installation but rather large for portable operation (It won’t fit in the car!). I’ve built a smaller one for portable (or mobile?) with 0.75m sides, using the same construction method.
As expected the Mk.II is 7.5dB down on its larger brother. It is still sensitive enough to detect external noise but requires some additional amplification. Whilst my IC-703 and TS-870S receivers have switchable preamps to make up this deficit, my home-brew receiver feeds the signal directly into a Plessey SL6440 mixer IC, so some extra gain is needed
I built a simple wideband RF amplifer using a design originally published in Solid State Design for the Radio Amateur, by W1FB. Some component values differ from the published design, mainly dictated by what I had available. The transistor used was a BFY90, the input, output and decoupling capacitors are 0.1 micro Farads. The gain was measured at 19dB between 0.2 and 100MHz.