A time domain reflectometer sends a pulse down a cable and times the echo, and the distance it gives you is arithmetic on the speed of light in that cable. It puts a fault on this pair at twelve hundred and forty metres. The cable is nine hundred and thirty metres long and ends in a sealed joint in a field.
I am a network engineer and I do overnight fault work on copper, which is a trade people assume has ended and has not.
How the instrument works, because everything here depends on trusting it.
A time domain reflectometer sends a fast pulse down a pair of wires. Wherever the impedance changes, some of that pulse comes back. A break reflects one way, a short reflects the other, and a joint or a wet section gives you its own signature.
You time the echo. The pulse travels at a fraction of the speed of light that depends on the cable's construction, which we call the velocity factor, and for the paper and polythene cables in this country it is known and published and on the tester's menu.
Time times velocity, divided by two because it went there and back. That is it. No interpretation.
In twenty two years I have put a spade in the ground where that instrument told me to, hundreds of times, and the fault has been there.
The pair I am talking about runs from a cabinet on the edge of a village out to a farm.
Nine hundred and thirty metres. I have the records, I have walked it, and I have seen both ends.
The tester puts an open circuit fault at twelve hundred and forty metres.
Three hundred and ten metres past the end of a cable that ends.
What has been checked and eliminated, and some of this I did the first night and some took four years.
The velocity factor. Wrong factor, wrong distance, and it is the mistake a trainee makes. Our records say what that cable is, and I have confirmed it by testing a known length of the same cable from the same cabinet to a pole I can touch, and the tester reads the distance to the pole within four metres.
The instrument. Three testers, two makes, one of them brand new out of the stores in 2023.
The far end. The cable terminates in a sealed underground joint in the corner of a field, which in that design is where the overhead drop to the farm starts. I have had that joint up twice, in daylight, with a digger and a jointer.
It is a proper joint, dry, with the right number of pairs in it, and the pair we are discussing is jointed through to a drop wire that goes up a pole and into the farmhouse.
And past the joint there is nothing. The cable stops. There is no continuation, no old route, nothing on the plans, and the field behind it is a field, and I have walked it with a cable locator on the most sensitive setting and got nothing but a water main.
The detail that makes this my problem rather than a curiosity.
The line works.
The farm has a working service on that pair. It has dial tone, it carries their broadband at a rate consistent with nine hundred and thirty metres of copper, and it has not had a genuine fault reported since 2016.
So the pair is continuous and healthy to nine hundred and thirty metres, where it is jointed, and beyond that point it continues for three hundred and ten metres to a break.
Both of those are measurements and they are taken with the same instrument within about a minute of each other.
What I did in 2022, which I have not told anybody at work.
A reflectometer shows you a trace. Distance along the bottom, reflection up the side, and an engineer reads the shape.
I saved the trace. I have eleven of them now, from four testers, and they are on a laptop in a folder.
Past the joint at nine hundred and thirty, which shows as the little blip a good joint makes, the trace is flat. A cable that has ended gives you the big clean reflection of an open circuit right there at the end.
This one carries on flat, the way a continuous cable does, for three hundred and ten metres, and then opens.
And in the flat part there are two small blips, at about a thousand and at about eleven hundred, and they have the shape and the size of joints.
Somebody with a tester and no other information would read that trace and tell you there is another three hundred metres of cable out there with two joints in it, and would ask why it is not on the records.
What I do about it.
The pair is marked on the system as tested good to the joint, with a note that there is a reflectometer anomaly beyond the termination and that it does not affect service.
That note is mine and it has been there since 2019 and nobody has ever asked me about it.
I have not requested a dig. I thought about it for a long time, and the reason I have not is that a dig needs a reason on a form, and the reason would have to be written by me, and the honest version of it gets my competence looked at rather than the field.
The only thing that has changed is when I test it.
It is on a routine and it comes up about twice a year, and it used to be a night job like everything else, because that is when you can take a line down.
I do it at the cabinet in daylight now, on an arranged interruption, with the farm told in advance.
In February 2019 I tested it at about two in the morning, parked on the verge with the cabinet open and the laptop on the roof of the van, and the field behind the hedge is on a slight rise.
The trace came up as it always does. I looked at it for a while, and then I looked up, because the reason I had gone out at that hour in the first place is that the fault was showing as intermittent that week.
It had gone. The trace was clean to nine hundred and thirty and opened at the joint like a cable that ends.
I took a reading every five minutes for an hour and at eleven minutes past three it came back, all three hundred and ten metres of it, the two joints in the same places, and I packed the laptop in the van and sat in the driver's seat with the door shut until the light came.