I hope it helps you! Report back if you see an improvement. 🙂

Since recording that, I’ve made a test rig kit available: electronics.halibut.com/product/common-mode-current-choke-test-rig/ I’m not sure how good it’ll be at VHF and UHF, but it should work well enough at HF. For VHF and up, I’d made individual adaptors for each connector type, and keep all wire lengths as short as possible. I also wouldn’t bother with switching.

That method certainly works, but it’s effectiveness is linear with the number of beads you put on the coax. Wrapping coax around the toroid or coiling it, the effectiveness goes up with the square of the number of turns. This only means it takes a lot of ferrite material (read: cost) to have the same rejection. It still works the same. Having said that, the linear version doesn’t increase capacitance, so you don’t have the self resonance problem I mentioned in the video. That’s a benefit to doing it the way MFJ does it. Isolator vs choke: yeah, I think they’re basically the same. Assuming your description of the 915 is right (just a length of coax with ferrite beads on it), it’s a common mode current choke.

Three turns, two clamp ons: It would not be exactly the same, no. I don't have measurement data at hand, but my memory was that two separate 3-turn loops, each with one clamp-on, provided better rejection than two clamp-ons on a single 3-turn loop. The clamp-ons increase the inductance of a loop without increasing the capacitance, which is good. But the loops are the root of the inductance; having two of them in series is better than having one. Three ferrites vs three turns through a single ferrite: These are VERY different. Adding ferrites to a cable is linear, but adding turns to a coil is geometric (goes up with the square of the number of turns.) Think of a single ferrite on a cable with no loops as a 1-turn inductor (which is what it actually is). When you put three ferrites on that cable in a line, that's three 1-turn inductors in series. When you put inductors in series, their inductance adds. So the inductance is i*n (i being the inductance of a 1-turn inductor, n being the number of ferrites). However, adding more turns to your inductor causes the inductance to go up with the SQUARE of the number of turns: i*(n^2). So, this is definitely a more efficient use of expensive ferrite material. There are draw-backs to adding more turns, however: 1) It is physically larger and possibly more awkward than just a straight linear string of beads on a cable. Stringing beads on the cable may be the more PHYSICALLY preferred solution. 2) As shown in the video, more turns also increases the mutual capacitance between turns, which makes the choke more resonant, and results in that very peak-y graph. So more turns is not always better. That's why I suggested sticking to three turns, and adding more inductors in series. Does that answer your questions?

Yes, exactly that. My tests of the Gordian Knot chokes shows why you absolutely want to keep the turns from overlapping. I'm really bummed I didn't make a proper ugly-balun to include here. I actually had a slide at the end that I cut out because I didn't do it justice.

I have not! Halibut Electronics will start selling a test rig before too long (give me another couple months to finish getting the business setup). If you've got that Unun, I'd love for you to test it and post your results. :-)

Also: thank you for the kind words! I’m glad you enjoyed the video and found it informative. :-)

Ignore the post I just deleted, I misunderstood your comment. You’re talking about the ground link between the two sides. Uhhh…. No, it’s not. Good catch. *blush* They are connected back at the VNA itself. So the current has a path. It’s just not the path I’d prefer it to take. The fact that I kept the jumpers between the VNA and test rig short May be what saves me here. But yeah, that’s not good. Thank you for catching that.

@@SmittyHalibut The pleasure is all mine. Good luck with 'secret squirrel'