I would have to disagree with Martin about coupling loop size re efficiency.
I find it the opposite. More efficient= bigger feed loop.
First the 1/5 rule is a staring point, and really should be 1/4..its easier to make smaller than bigger!
To quantify my results on feed loops..
I have currently two efficient large mag loops..vac cap, copper tube etc..
One single turn for 40-60m and the other twin turn 80-160m.
Obviously both are more efficient on the higher frequencies ( ie 40 and 80.)
The feed loops for both are rotatable on the main loop axis, ie i can swing them from inline with main loop to 90 away.
Both require max coupling ( inline with main loop) for higher frequencies ( most efficient operation of loop) and less coupling at lower frequencies.
Typically this on my loops equates to about 1/4 size on higher to 1/5 on lower.
There of course maybe something else at play here too..
Our good friend Leigh also states ( see Steve¡¯s vk5sfa webpages) to make coupling loop bigger than 1/5. Steve¡¯s loops are efficient no compromise loops designed with Leigh and setup with Leigh. ( my twin turn loop is based on this article.)
Simon G0zen