Disclaimer
No claim of accuracy is made for this calculator. Do not use this calculator for any purpose except entertainment. Refer to the references and perform your own calculations to see if they agree with this calculator's output.
Notes on usage of this calculator
This calculator allows calculation and visualisation of loss components in a small transmitting loop. Unlike other small loop calculators, this loop calculator has a number of unusual features:
- Support for multi-turn loops (but see notes below)
- Choice of copper, aluminum, or custom conductor material
- Support for flat (rectangular cross-section) conductors using W9CF's analytic formulas for RF resistance and self-impedance (see references)
- Support for braided conductor loss using a fixed multiplier between 4 and 16 to represent increased losses through wire-to-wire pressure connections in braid
- Inclusion of capacitor loss including fixed, dielectric, and metal losses, as per G3RBJ's model (see references); two capacitor models pre-defined, with custom parameters allowed
- Inclusion of connecting wire loss (connecting capacitor and loop)
- Visualisation of all loss components in both normalised and absolute-value graphs.
Note that for multi-turn loops, the calculator does not include losses due to turn-to-turn proximity effect or turn-to-turn capacitance (circulating currents). The efficiency estimates for multi-turn loops will thus be optimistic.
The validity of W9CF's analytic formulas for rectangular cross-section conductors as used in small transmitting loops, especially multi-turn loops, is not clear.
Ground losses are not included.
I tested the calculator with some values for a QW (1/4 WL) circumference loop, the Rrad value is off (low, as in factor-of-ten kind of 'low') like other loop calculators. This aspect REALLY needs to be addressed by somebody else (besides me and Ben Edginton G0CWT) and noted in the 'Notes' or caveat section of the instructions ...
返信削除73, Jim
Thanks for taking the time to test and comment on my loop calculator. I appreciate the feedback.
削除I am familiar with G0CWT's feed method for quarter-wavelength loops (http://www.g0cwt.co.uk/magloops/new_page_6.htm ). I think G0CWT is mixing together two ideas (feedpoint resistance and radiation resistance) which leads to erroneous conclusions. Quote by G0CWT: "the ARRL Antenna book which told me that the radiation resistance of the loop would be less than 0.1ohm on 80 metres and I took this to be the feed point resistance. [But] the feed-point resistance of a quarter wave loop next to the capacitor is 22.3 ohms." Finally, he claims on page http://www.g0cwt.co.uk/magloops/new_page_4.htm the following. Quote by G0CWT: "This means that because of the relatively high impedance at this feed point losses from dc resistance in the loop are at a minimum and thick tubing is not necessary."
The problem is that this 22.3 ohms feedpoint resistance is not the radiation resistance. Merely changing the feedpoint does not change the ohmic losses (AC resistance) or the radiation resistance of the loop.
This was discussed here: http://www.eham.net/ehamforum/smf/index.php?/topic,66696.msg436088.html#msg436088 . In that thread, N3OX explains it best. Quote by N3OX: "Moving the feedpoint in a way that doesn't change the current distribution on an antenna doesn't change its efficiency. So if you would agree that the loop has 1 ohm radiation resistance and a few ohms total resistance when fed opposite the capacitor, that tells you the efficiency. The high impedance feedpoint is convenient but does *not* change the basic ohmic losses in the antenna. [...] John Kraus's formula for radiation resistance comes from an analytic analysis of the current on a small loop. It's does start to depart for larger loops because of an assumption made to do the integral, but it doesn't depart much. EZNEC and 4NEC2 are just solving Maxwell's equations. [...] That 22.2 ohms is a combination of radiation resistance and ohmic loss resistance, transformed up to a higher value than what you would find at the point opposite the capacitor. It is NOT the radiation resistance. In my opinion, there's confusion on G0CWT's part regarding what radiation resistance means and what it is for various sizes of loop antennas. It is not teeny tiny for 1/4 wave loops, that is true. But it's not 22 ohms either. The resistive losses for the wire loop are already contained in that 22 ohm value, but in a somewhat non-obvious way. I am not convinced by G0CWT's website that the losses are negligible..."
I'm sure that 0.25-wavelength-circumference loops work significantly better than 0.10-wavelength-circumference loops. The radiation resistance does increase when going from 0.10-wavelength-circumference to 0.25-wavelength-circumference. Traditional theory and NEC-2 predict this. But G0CWT's claims go a step further and imply that the radiation resistance (and efficiency) improve further just by moving the feedpoint. I haven't seen sufficient evidence by G0CWT to support claims of higher efficiency. His reports on http://www.g0cwt.co.uk/magloops/new_page_4.htm do not include objective measurements but only the fact that he was able to make contacts. A proper test would require careful and objective field strength measurements against a reference antenna. One way to do this is excellently illustrated by K4HKX here: http://www.qrz.com/db/K4HKX . K4HKX uses hundreds of averaged, objective, and automatic signal strength reports from the Reverse Beacon Network (http://www.reversebeacon.net/) to compare performance of different antennas.
I'm open to consider any new theories that are backed by evidence. But right now, I don't see sufficient evidence that supports G0CWT's theory of improved efficiency just by moving the feedpoint. I think traditional theory and NEC-2 can already explain the good performance of quarter-wavelength loops.
Thanks again for your feedback!
-qrp.gaijin
re: "The problem is that this 22.3 ohms feedpoint resistance is not the radiation resistance. Merely changing the feedpoint does not change the ohmic losses (AC resistance) or the radiation resistance of the loop."
削除Calculate the AC resistance of the wire involved. It is a LOT less than 22 Ohms, so you are wrong on this point, and the 22.5 Ohms IS the Rrad (radiation resistance.) This is easily has been easily proven here, just now.
Use AC resistance tool here to calculate AC resistance of wire used:
https://thetruthaboutloops.wordpress.com/tools/
To summarize, again, loop calculation formulas for 1/4 Lambda circumference loops are incorrect. I am using successfully 1/4 WL loop on 160 meters, about 110 feet circumference in a rectangular shape, 16 ft x 40 feet vertically oriented.
返信削除Also, NEC2 makes same calculation errors for computing Rrad.
I even fed loop 20 ft to one side of tuning cap and feedpoint Z / the Rrad is still around 20 Ohms.
I now use a split-cap feed network, like Patterson Match on loop antenna, no more magnetics, no more ferrite transformer.
73, Jim (Running 160 meter WSPR tonight, July 10 2024 from 5 call district in TX)