Connector-less connections?

I can imagine testing the whole thing, or even individual wires, with capacitive coupling through the insulation.

TDR or capacitive measurements would tell you a lot from one end.

But is a cable likely to be damaged somewhere between the ends?

Use wi-fi!

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

This is a classic problem. The classic solution is conduit, which can be plastic for low-power signal cables. (For extra shielding, use grounded steel conduit.)

Grunt labor installs the conduits (including a pull tape) and closes the walls. Better grade of grunt labor pulls cables (plus a few extras) into the conduits.

Then test and connect the cables, using only cables that work.

Joe

I wonder if there's an IDC fixture for standard cat5e/etc type cables, without having to untwist the twisted pairs?

ie the installer feeds through the brown pair through the brown slot, blue through blue slot, etc. Then squeeze the two parts together and the IDC grabs the conductors, making contact and connects to an RJ45 socket.

You end up with something where you potentially have swaps within pairs. Maybe you don't actually care about that if both ends are swapped (the signal doesn't care the colour of the insulation), or maybe that's fixable later with jumpers on the PCB.

Then this is both your test fixture and final termination.

I don't think a typical ribbon cable IDC would do it, but perhaps there is another sort of IDC socket which is set up so a twisted pair would get clamped and pierced correctly?

Theo

There is twisted-pair cable that has ordinary flat sections at regular intervals.

Jeroen Belleman

These perhaps?

I've not seen one. There are some 8P8C's that allow you to push the (set of) conductors clear through the connector shell, exiting at the leading edge of the connector shell (where the contacts are located). After crimping, you trim off the excess with a pair of side cutters.

These are nice in that you aren't trying to keep all of the conductors firmly engaged in the shell while crimping. Annoying to jam the conductors in, crimp and then discover one was "too short".

Even nicer if you don't care which conductor ends up where! :>

Note I wouldn't care if the brown pair ended up in the blue slot... or, even if the brown pair got *split* and ended up SHARING the blue slot with the blue pair. The test set could sort out which wire is where before checking the cable.

And, as the installer is going to cut off that connector, there is no need for *it* to maintain any "standards compliance".

See above.

I can't use 8P8C's. They are too "massive" (amusing choice of word for such a tiny part). But, when the only things at the endpoints are service loops of cable, I can tolerate the presence of the connector; tuck the end of the tested cable into the jbox and attach cover plate. Installer will, at some future time/date, open box, remove connector and install endpoint device(s).

If the "wire puller" (grunt) only has to affix a connector without concern for which wire goes where, he can use a "regular" 8P8C without taking care to arrange the individual conductors correctly. He'd have to ensure each actually made contact (hence the sort of "push through" connector shell mentioned above) and made a genuine effort not to lose the twists in the pairs.

The downside is you have to discard a pair of connectors for each cable.

And, *fit* that pair before you can test.

Imagine you can fit and crimp a connector in a minute (remember, you're on a ladder, in a crawl space, etc. while doing so, not seated at a bench). That's 10 hours of work for 300 drops;

Not counting the time it takes you to pull or move from point A to point B.

Hence the (ideal) appeal of just cutting the cable and jamming the "unprepared" end into a test set; no need to cut, remove jacket, separate conductors, crimp connector, etc.

Barring that, being able to remove the jacket, splay the individual conductors unceremoniously into a RUSABLE connector IN the test set and qualify the cable. Disengage conductors and move on to next cable.

Note that you need to verify each conductor, not just "mechanical continuity" for the cable itself; a conductor may have been damaged during the pull (the contractor may have had a *carpenter* pulling the cables -- it's just a thick string, right? <rolls eyes>)

Ignore the extra material and labor cost (for the moment).

Problem is one of scale. Walk through YOUR house. Count the number of duplex receptacles, individual light switches (not the number of BOXES with switches in them but the actual number of switches), light fixtures, other electronic devices (tstats, doorbell annunciators), CATV/phone connections, etc. Imagine EACH is fed by an individual cable that can be traced back to the main panel (network switch, in my case).

Chances are, you won't come close to the numbers involved.

And, you'd have to run a conduit for each of those individual drops. You end up with a shitload of intermediary pull boxes, often in close proximity to each other.

You can only get a few (4) cables in a 1/2" EMT (more in larger conduits) and typically can't run a "farther" cable through a "nearer" box due to the locations of those boxes. This differs from mains wiring where you can "start" the run to the next box from an upstream box.

It's easier to just add loom to cable "bundles" after the fact (to dress it up and provide increased isolation from other cables)

And, it still does nothing for TESTING the cables.

The connector, itself, is HUGE! :<

But, treating the "clamshell" as part of the test set, the "core carrier" (their term, not mine) MIGHT be reusable. It seems to suggest the insulation is pierced when the clamshell closes on it.

No doubt it (core carrier) can be extracted from the clamshell, later.

But, I'm not sure if the pierced conductors can be removed from the "core carrier" once it has been engaged.

yeah, but if one of the cables is damaged why is that your cost?

why can't it stay?

anyway

Perhaps install the cable without cutting it - then you only need to terminate two ends to test it.

Solved problem. Use larger conduit, and have "data closets" in conveniently-located closets. I'm describing a standard way that office buildings are wired.

Joe

Office buildings, plants, etc. have far more places to hide that sort of thing than most residences. And, are more tolerant of its presence/appearance.