Randy, I do not use crimp tubes at all. My components are, from one end to the other, servo arm, sullivan clevis in servo arm, threaded coupler threaded into clevis, kevlar thread CA'ed into the hollow tube of the threaded coupler. That's the servo end. In the middle I like to run some kind of nyrod from an inch or two behind the servo in a fairly straight line to the fuselage exit, where I trim the nyrod flush with the fuselage. I epoxy or goop the nyrod at servo end and tail end, and also somewhere in between depending on what structure is available to goop to. I glue a 1/8" thick balsa scab inside the fuselage at the nyrod exit, just to have more of a glue joint. When I say "fairly straight line" it's not geometrically precise, but I just don't want much of an angle where the kevlar exits the nyrod on the way out to the rudder/elevator. It is the one or two inches of kevlar at the nyrod exit that I don't want to expose to fraying, so I CA that part of the kevlar thread to protect the fibers, and I run the CA all the way to the rudder/elevator end, so the "thread" is now stiff and will slip into the hollow tube of the threaded coupler easily.

The CA'ing of the thread is done to measure before I hook up the terminations on either end. I just clamp a length of thread from whatever skyhook is handy, and clamp a piece of something heavy to make it hang down straight. Then I CA the servo end for a few inches, and the rudder/elevator end for whatever distance will run from inside the nyrod to the rudder/elevator coupler.

When the thread is cut to approximate length and CA'ed on both ends, I do the servo end connection: CA the kevlar inside the threaded coupler (no crimp tubes anywhere), thread the coupler into the clevis about halfway in, snap the clevis into the servo arm. Then I slide the other end of the kevlar into the nyrod as far as it wants to go. Sometimes if I hold the fuselage nose up the kevlar will slide right out the other end; sometimes it hangs up, and I just kiss the exit and suck air through the straw, and that pulls the kevlar through. Now I have 6-8" kevlar thread hanging out the end. I clamp my rudder/elevator to neutral, snap the clevis onto the control horn, slip the kevlar into the threaded coupler and hold it up so it's in line with the clevis. I set the end of the coupler even with the middle of the threaded barrel of the clevis, pinch the thread and coupler together at that point, and touch a drop of CA into the nearer end of the threaded coupler. Most of the time the CA wicks through, and I find my fingers are glued to the other end of the coupler. I'm used to that, so I just pull the thing off, leave some of my dead skin on the coupler, and for good measure maybe put another drop on the farther end of the coupler. I use a light pinch when I do this, so if the CA comes all the way through to my fingers, I don't have much of a grip on the threads of the coupler. I don't want CA fouling the threaded surface. It really helps to use a micro-thin tube for your CA to keep from dropping too big a drop into the coupler.

Now I trim whatever excess length there is of the kevlar, and thread the coupler into the clevis. You now have maybe 1/4" of exposed thread on the couplers to take up on either end, so you can take up 1/2" of slack in the line, if there is that much. Usually there isn't. If that's not quite enough, you can twist the kevlar thread, and that will take up a bit more slack.

The kevlar "cable" does not need to be guitar string tight, all it needs is to have zero slack. When I have both sides rigged, I turn on the tx and play with the stick while holding the rudder/elevator lightly, just to see how loose/tight I am. It needs only light tension, but it should not be loose. I aim for being able to deflect the rudder slightly--maybe 1/8" or so-- before the servo arm will move.

I have learned to put a 2-56 locknut on the coupler before I thread it into the clevis, so when I have the length and tension just right on both sides I can lock it in. It is important to have pretty much the same tension on either side of a pull/pull cable pair.

There is a problem which has not (so far) troubled me, but I have seen an installation where, when you moved the stick from one side to the other across the neutral point, the rudder did not transition smoothly, but at a certain point seemed to hop or snap from one side to the other. That I suppose is a hinging problem, and if I had that problem I think I'd probably re-build or at least re-hinge my rudder.

Just to provide a solid bearing for whatever tension there is in the pull/pull rigging, I like to have a pinned hinge as close as possible to where my control horn is.

As to knots, I just like CA better. You can CA to a fairly precise length with good control on tension. I have been a fisherman since I was a pup, and think I'm adept at knots, but haven't seen any that I like as well as this CA technique.

Fishing for bluefish, tuna, sharks, etc. I have rigged a lot of nylon coated steel wire leader with the crimp tubes and all that, so my preference for kevlar is just that--a preference. I know multistrand steel wire is used by a lot of better fliers than I ever hope to be, so I don't knock what works for the other guy. I hope this is helpful.