Here come the numbers... (bonk bonk bonk!!! for the insiders):
- radiator total length: 327cm
- matching stub length: 46cm
For the radiator, i.e. loop, this is the actual wire cut length. Meaning, there's about 5mm vanishing into the luster terminal.
As for the stub, same story, however, the stub is shorted, and whatever is needed for that short is not taken into account by that figure. However, I believe, that this would not make a big deal difference in matching anyway...
This is, how stuff is wired up in my (experimental) setup:
The "matching stub" (current center of the loop) is connected to both terminals, either of the loop leads to the respective terminals. The whole thing is matched to RG58 by means of a 4:1 balun using a T50-6 toroid contraption (classic approach).
With this setup my MFJ-analyzer was showing a SWR of 1.2:1 @ 50Ohms. Fine with me. The loop is sensitive to its environment however.
Short note on how I resonated this loop. It was cut to a frequency much lower than wanted, since my aim was a single frequency loop for a grabber. Anyway, the 30m band is not that wide, or is it? With the analyzer engaged, I progressively cut off bits of the "voltage"-end of the loop, until I reached the aimed mid resonance frequency. This should be done symmetrically, in order to prevent noise caused by local E-fields (vacuum cleaners, hair dryers, etc.).
No toroid in the junk box? This could be an alternative to match TV-twin lead to 50Ohmx coax:
It was giving a slightly inferior result, hence, the T50-6 balun won and will stay connected with the grabber for the upcoming future.