• So I guess my question is, at what tuning stage do I have to rotate the loops to set insertion loss.

?

That’s the very first thing you do.?

?

From the factory, the loops would have been adjusted such that, at your 600 kHz pass-to-notch split, each cavity has 0.6 dB of insertion loss.? So, measure the insertion loss at the pass frequency with the notch set somewhere around 600 kHz away from the notch (exact frequencies are unimportant), rotate the loop until you get exactxly 0.6 dB insertion loss (the pass-notch delta will change as you adjust the loops so you may need to tweak the notch again to maintain roughly 600 kHz offset), lather, rinse, repeat, lock it down.? Do the same for all six cavities.? You can do this with the Nano VNA or a service monitor with spectrum analyzer and tracking generator but calibrate out all cable losses so you’re sure you’re right on 0.6 dB insertion loss.? You are NOT adjusting anything for return loss in this step, nor do you care about exact frequencies of the pass and notch, only that they are kept nominally 600 kHz apart during adjustments.

?

THEN tune each cavity individually for best return loss at their respective pass frequency using a standard tee and 50 ohm load installed on the top of the cavity.? Your Nano VNA or a good return loss bridge with the SA/TG is the tool for the job.? The tee and the 50 ohm load you put on the tee must be high-quality; don’t’ use a run of the mill dummy load or even a service monitor antenna or T/R input as the termination.

?

While still unharnessed, set the notch frequency for each cavity individually but do NOT touch the pass.? The pass tuning will not change as you fine-tune the notches provided that you had kept them somewhere near 600 kHz offset during the original loop rotation step.? Again, you can use the Nano VNA if it has enough dynamic range, or use the SA/TG (without the return loss bridge obviously).? The cavities on the high-pass side should end up having the same notch depth (within +/-1 dB), and the three on low-pass side should have the same notch depth, but the two sides may not be equal.? That is, the high-pass average might be 40 dB per cavity and the low-pass average might be 38 dB per cavity, but the cavity-to-cavity variation on each side should be within 1 dB assuming you set the coupling identical on all in the first step.

?

Then harness everything back together.? You should NOT need to touch the pass tuning, and certainly not the loop coupling, beyond this point.? However, because you replaced your test tees with the real in-line harness tees, the notches will have shifted slightly.? Touch up the notch tuning if it doesn’t meet spec or if it’s obvious that, when sweeping with instrumentation that has sufficient dynamic range, you can actually see that the notches aren’t spot-on.? This is not something you’re going to be able to do with a Nano VNA, nor even a SA/TG unless you have some external amplification.? However, there are ways you can cheat by connecting only two cavities at a time (i.e. eliminate one of the notches from the sum-total so you can see the two that are connected) – search the archives if it’s not clear how to do this as I’ve written it up before.

?

??????????????????????????????????????????????????????????????????????????????????????????????? --- Jeff WN3A

?