I am working on putting together a 40 meter delta loop using a 1/4 wavelength of RG-6 as the matching section. However I'm unsure of where to create my feedpoint choke.
Do I coil up the RG6 right at the feedpoint of the antenna itself? Or do I coil up the 50 ohm RG8x at the point where it feeds into the 75 ohm matching section?
EDIT: My choke consists of 8 turns of coax through a mix 31 big clamp on bead.
The choke should go at the transition between unbalanced and balanced, which is where the RG-6 meets the ends of the loop wire.
You talk of coiling coax: I suggest you coil it around a ferrite core and not just air. The trouble with air-core "chokes" is they have no resistance, only reactance, in common-mode. Depending on what the common-mode impedance is in the rest of the antenna system, adding this reactance might actually tune the common-mode to resonance, making things worse. A good choke has a substantially resistive impedance: this is guaranteed to raise the overall common-mode impedance in all cases. See additional explanation by G3TXQ.
Forget about air-core chokes. G3TXQ's page that Phil links to in his excellent answer here shows just how much superior ferrite chokes are than air core choke baluns.
Please visit k9yc.com and download rfi-ham.pdf. Jim Brown there on page 36 recommends a choke balun made by wrapping 8 turns of the coax at the antenna feedpoint through four stacked 2.4" diameter #31 ferrite cores. They cost about $7 from http://www.mouser.com. And don't wrap the RG-6 too tight! That PDF tells how to properly wrap it.
Below is a similar choke balun that I made, at the base of my 160m inverted-L. Those are the exact Fair-rite #31 cores I mentioned above, with RG-6 coax. You don't need to make the loops as big as I did.
(For 1.8 MHz, it really should have had more turns and cores, but it worked very well nonetheless.) I used this antenna on 160 through 20 meters, with great results.