Is this what you meant? http://i.imgur.com/tqEbC1A.png
The numbers indicate the turn number, so the black line turns into the brown line underneath the cruciform.
Bloop, you’re on the right track. And, you’ll need to try
various locks with the particular conditions you have
–cord & object surface. But note that in the image
you give (numbered 1-2-3 with black-orange-green-blue)
the black line (would be #0 :o) is UNsecured, and so
this doesn’t meet your need for security when slack.
What I described was to have #2 cross over #0 (black)
and under #1 after 1 crossed under 0, as a means to
secure #0. It’s a little tricky given the way in which #0 arrives at the knotting surface, and with the flat
surface itself.
Another idea is to take #1 away from its crossing at
the top surface to pass over #0 in turning around the
near side face and going back around, to come at the
position of “3” (well, snugly against itself on that side)
and then be tucked over-&-beneath itself (where it has
just crossed over #0, and so might secure it). This would
make an overhand (binder) knot in the orange part.
How well this works will depend on the materials.
Note that it’s nice in that each successive step doesn’t
require some prior anticipation of the step --what’s done
is done, and added to, not needing to anticipate the later
step.
It might be that a quick & effective securing could be made
by first wrapping #1 twice OVER #0 (so, well securing that,
and the friction of the wraps, held in hand, holding the desired
tension as you go to lock by …) --and the slope of the underside
ridge should help wraps press towards the tensioned line’s entry
there–
… THEN tucking the 3rd wrap under itself (but going from left
to right, opposite to the tucks you show)
and then back around and under as for #3 under #2.
(And hoping that the coating of the cable is relatively frictive
and gripping vs. slipping!)
Not sure where it might be best to PLACE the tucks;
you can try the top or the side (and maybe first one
then finish at the other place). If the overwraps done
first have decent effect at securing the line out of the
hole, the first tuck might be best done on the side,
so as not to threaten the pressure of these wraps?!
You can modify roo’s idea and avoid threading the cruciform bores. It would consist of a length adjustable 2-part threaded sleeve affair placed between the cruciform and your knot. The adjustable sleeves would not grip but simply float on the cable. You would adjust the sleeves to their shortest length setting, tie your knot, and then adjust the sleeves to the desired length for proper tension. This method eliminates the finicky nature of tying the knot under tension and allows for precision adjustment at any time due to knot slippage or other. If the knot tended to slip (requiring too frequent re-adjustment) this problem could be solved if the finished knot sat in a well of appropriate size - whose walls would need to constrict the knot to some degree - and preclude having to glue knot as a last resort. If gluing the knot was found to be necessary, the well would act as container to glue the knot into.
Maybe simpler yet: slip a small open coil compression spring - of appropriate size and strength - over the end of your cable and then a washer (preferably welled) with a hole just large enough for your cable to pass through. Then, when you tighten the knot, the spring will keep the desired tension on the cable.
I like the concept of the spring. Look, if this wire has no give like basically all wires have no give, then this wire is going to be rather tight or just loose. There is no in between. There’s no use in trying to make things as tight as possible because a couple of millimeters of slip would make your wire loose again. A spring would install some give to fool the wire into believing it’s a rope so that the wire can be an overachiever.
Anyway, bloooooop, do you find the wire has no give to it? Or am I off?
I initially used fishing line which had some give, but went to wire because of its lack of give. In the base, the pulleys have a non-circular profile, which only works if the line has no give at all. If there is any give in the wire system, the wires on the opposite side go very slack and fall out of the pulley groove.