I use Glenn's, KI6GD's, excellent "Magnetic Loop Calculator" (I have v1.4, 2009) to help in the design and experimentation with small loops. For 20m, here are the numbers (the program works with either Imperial or Metric units and various materials and shapes (copper, aluminum and shapes round, square and octagon) all which will affect the final product. Note: According to my ARRL Handbook, Belden RG-213 has a rated breakdown voltage of 3.7kV (RMS) and at the 50 watts operating power level the capacitor voltage is 3.3kV. That's cutting the safety margin close, so I wouldn't push the loop past 50 watts and most my operating is 25 watts and way less. But, just how good is any of that cheap, no-brand "Made in China" coax, eh?
Efficiency is around 53% for 20m which seems to work since I can work DX and they can work me! When used on 30m it's just over 26% according to the calculator and I've only been able to work Canada/U.S. stations, mostly. A larger loop would be a better choice for 30m to increase the efficiency.
Loop Circumference 3.03 m (10' loop circumference, circular shape)
Conductor Diameter 12.7 mm (1/2")
Frequency 14.097 MHz
Power 50.00 watts (for maximum capacitor voltage calculation)
Bandwidth 16.9 kHz
Capacitor Value 33.5 pF
Capacitor Voltage 3.3 kV
Conductor Wavelength 0.151 lamda
Efficiency 53.1 % (circular is the most efficient, followed by octagon at 50.4% , then square at 41.1%)
Inductance 2.995 µH
Inductive Reactance 265.4 ohms
Loop Area 0.7 m² (8.0 feet²)
Loop Diameter 1.0 m (3.2')
Loop Q Value 832.4 Qres (outrageously high Q value at resonance, thus a narrow bandwidth but less noise with the signal)
Radiation Resistance 0.085 ohms
Resistance Loss 0.075 ohms
A few years ago, I wrote a couple of articles on modifying and using WA8LMF's 30m magnetic loop design. I've re-edited and put them on my website. If anyone is interested, you can download the ZIP file from here: http://tinyurl.com/cevk44m.
This is the copper tubing magnetic loop that I'm testing outside. Since it has a stand, it's easier to set or move around and move stable in wind over the MFJ-933 tuner/wire loop. The 12.5mm (1/2") diameter copper tubing is wrapped in high-voltage rubber tape (60 kV) just in case someone were to touch the loop when it was transmitting! The copper look is 3.03m (10') and in inner, inductively coupled drive loop is 0.606m (2').
Tuning for 30m to 10m is via RG-213 coax stubs (outer jacket and braid removed). I used a Marette connector (wire nut to non-Canadian) threaded into one end of the copper tubing so the stubs can be changed. They just slide into the other end of the tubing and I use an MFJ antenna analyzer to adjust the resonant frequency by adjusting how far the stub slides into the tubing. Since small loops are very shape in tuning, it's often just by millimetres!
The basic design for the loop came from Stephen, WA8LMF, http://wa8lmf.net/ham/30m-magloop-ant.htm, and I made slight changes so it would work multi-band with stubs you can swap in and out. The loop works very well from my WSPR tests. Even at 500mW and 50 percent loop efficiency (20m and 90 percent for 10m), I'm hearing European stations and they are hearing me. At the 200mW level I'm only heard on this side of the Pond. Still remarkable because compared to my outdoor wire antenna 50m Delta loop, the small magnetic loop is producing the same WSPR tx/rx pattern! So, small loops can and do work, if properly designed.
As it's sitting so low to the ground, the loop's orientation doesn't matter as it seems to be omni directional. In the picture, east/west is broadside to the loop. The yellow, collapsible work stand came from an outdoor halogen light stand. The legs can go flat against the ground and a sandbag or some other weight can be used to hold it down in the wind.
Well, finally got my Flex-1500, VAC and WSPR software working togetther (TX/RX/PTT and CAT). Had to setup the CAT as a Kenwood TS50 to get that to work.
Amazingly, my indoor magnetic loop antenna is doing a very good job for TX/RX with the 4m of #10 wire and MFJ-933. Will have to try the home made 12.5mm 3m copper tubing and coax capacitor tuning stub to see which one performs better. Since both are side by side, it's easy to switch between the two and make fairly comparative tests.
The Flex, itself is very accurate right out of the box (better than my Field) but I'll run the WSPR frequency calibration tests to see if I can get it even tighter.
Regardless, the statistical medium of my TX frequency is always close to the mark, even if some WSPR stations are reporting way above or below my TX frequency (for unknown reasons).