I've made many of these and have some comments. Hopefully without going too far into it, but sticking to the basics:
What makes this different from just a bare-wire loop attached to a transformer is that the split-shield coax loop has built-in common-mode rejection, helping to make sure you are actually listening to the loop, and not your coax run to it. But not from any magic shielding, but simply the way the loop is constructed.
How? The "shielded" part of the loop is the inner transmission line. The "antenna" part doing the work is the braid. The outside surface of the braid that is. To quell 20 years of thread hashing back and forth to prove it, do this: Put a snap-on ferrite choke, which attenuates common mode, or outer braid current, over one of the arms. Dead loop.
The feedpoint is the 1-inch gap. Being exactly 1-inch is not critical, but basically having some sort of gap is! Don't leave any frog-hairs of braid wire across it.
The best thing is that while the common-mode braid reception of the loop itself does all the work, it serves a dual purpose - keeping common mode shack noise from getting to the feedpoint!
It does this by working in reverse - shack noise travels along your coax run to the loop, and splits equally at the bottom going up both sides of the arms. Because of this equality of current, NO differential voltage from the shack induced noise is produced at the feedpoint gap. The same happens on receive - any signals arriving broadside on both sides of the loop arms at the same time, results in no differential voltage at the gap - the null basically.
Real world: Keep those coax braid cuts and lengths as symetrical as you can. Still, small differences in construction, ie slight construction imbalances, can lead to common-mode current differentials traveling down your shack coax to the feedpoint where you don't want it.
Whew - this is basically the loooong version of saying that I'd still recommend a 1:1 ferrite choke/sleeve on the coax just prior to the transformer.
Tip: I rarely used a preamp either, however I DO find that the use of a manual tuner helpful. You aren't trying to tune the loop to resonance, but instead reducing the reactance of the overall "system". There is a lot of capacitive reactance from the coax to tune out, despite the inductance of the loop itself. Tuning will typically be VERY sharp. If you do it by ear, do it at NIGHT where you can actually hear a difference. Of course an antenna analyzer makes very short work of this.
Because we are rx-only, don't sweat trying to obtain the perfect match / swr. Just do the best you can. Fortunately we're chasing the S/N, or signal to noise ratio, rather than overall signal strength.
