On Thursday, April 8, 2021 at 1:21:27 PM UTC-7, jbeattie wrote:
On Thursday, April 8, 2021 at 12:41:34 PM UTC-7, wrote:
On Thursday, April 8, 2021 at 12:16:39 PM UTC-7, jbeattie wrote:
On Thursday, April 8, 2021 at 10:49:48 AM UTC-7, wrote:
On Thu, 8 Apr 2021 05:43:43 -0700 (PDT), Tom Kunich
wrote:

Jeff, try picturing threading a Kevlar inner cable through the outer. Now how
would you expect to do that?
Two ways. Pull or push. For pull, I would run a metal wire through
the cable housing, attach to Kevlar fiber with glue, and pull on the
metal wire. For push, I would glue a wad of cotton to form a shuttle
on the end of the Kevlar fiber. A conical dart shape should work.
Insert the shuttle into the cable housing and blow it through housing
using compressed air.
https://en.wikipedia.org/wiki/Cable_jetting
As a real engineer I have some ideas that could work but why when
stainless steel is more than sufficient. Jay is a lawyer and not a
mechanic. We all know that he broke a cable probably because he
overtightened it at the derailleur and broke the strands.
We all don't know that. You might, but we don't.
Stainless is extremely good at weather resistance but very bad at
pressures that overload the molecular bond of the material.
Pressure? Stainless is excellent in compression (also known as
pressure) but might have problems in tension or torsion.
Pfff. I've been a bike mechanic since I had bikes, and I worked in a shop . . . and I had my own wheel building business in college . . . and Trek installed the cable. It was OE on my Emonda. And how do you over tighten an STI cable? It has to be adequately tensioned for index shifting. That makes no sense at all. The cable broke in the lever at the usual place -- about an inch behind the head, inside the lever. https://photos.app.goo.gl/Hoi29kauY5SW767m9 This should be where Tom criticizes the Japanese engineers who designed the 8000 series for not coming up with something that didn't involve winding-up a cable. There should be some hydraulic, linear-pull mechanism with digital controls and a fray-o-meter indicating when failure is imminent. Safety inflation!
You overtighten it by turning the locking screw too tightly or perhaps having a washer with a sharp edge on it. If it was OEM that means that it was probably improperly threaded through the locking mechanism down on the rear derailleur and simply took this long to break. None of those washers and stops is very intuitive and it is quite easy to run the cables through there incorrectly and put an un-designed for load on the cable. Perhaps they used 105 inner and outers which is 20% the quality of the Dura Ace cable set.

Like I say, I'm not knocking you because these things are easy to do if you aren't really careful. I only use Record cable sets and cut them ONLY with a Park cable clipper. I have used several other cable cutters and they ALL cut crooked. Especially on the Brake Outer Cables.
Did you look at the picture? The cable didn't break at the cable anchor (if that's what you mean by "locking screw"). There is no way that an installation error would cause a cable failure in the lever, except possibly lack of lubrication or maybe a bad cable guide in the lever, which I did not see. Who knows if the cable was sub-par, and there is no such thing as Shimano 105 inner cable -- although they have marketed some inner cable as Dura Ace. AFAIK, all the SS cables are the same but with different coatings, many of which become fuzzy with use. I stick with uncoated SS.

-- Jay Beattie.
Jay, I didn't see any picture showing the broken cable and must have missed that the cable broke at that damned sharp turn that they have on the latest Shimano levers. I wondered about that turn the first time I set a Shimano 105 up. There a much better designs that would completely avoid that turn altogether and have the shift cable go through the front of the lever similar to the brake cable. They even have everything in the assembly to use to do it.