(Edit - typed while MrSaffron and you exchanged posts about it).

The blue and blue/white wires are a pair twisted together. That helps cancel our randon noise attacking them as they both get the attack which self-cancels. A bit like (audible) noise-cancelling headphones.

The blue should go to 2, the blue/white to 5. Similarly with the orange and orange/white.

That picture shows that two pairs have been split.

When wired correctly noise is picked up by (say blue) T2 at one end to T2 at the other, and the same noise is picked up by the blue/white wire from T5-T5. Because they are wound round each other, it largely cancels out.

In the picture T2 will pick up noise on the blue and T5 noise on the orange, but they aren't wound round each other so don't cancel out. Similarly with the blue/white and orange/white. Both pairs as wired there will get additive noise instead of largely cancelling noise.

To us it seems a bit minor as they are in the same outer casing so getting similar noise exposure, but in practice the exposure is different enough to mess up the connection.

Edited by RobertoS (Sat 08-Oct-16 16:55:12)

As posted previously, this is not the case with my wiring. I am using the blue and blue/white wires from the same pair (Pair 1).

Yes, I saw that. Hence my edit at the top of my post . I thought it worth leaving the post there rather than deleting it though.

Is your CAT5 wiring solid copper, or is it stranded or copper-coated-magnesium, etc?

Proper solid copper. The same stuff used for commercial structured cabling.

I do not beleive the Cat5e is at fault as I get exactly the same results are the NTE5 test socket with everything else disconnected.

Be wary when estimating line length.

On a logical basis, my line would be about 50 Metres long, plenty of ducting available on a green-field site in 1967 - but is in fact at least 250 Metres, as it does not follow the "obvious" route, going about 125 Metres away from my house, before returning by about another 125 Metres.

Basically all is underground, compounded by the main ducting with the E-side from the Exchange, goes past my house about 2 metres away, to reach the PCP, 50 metres away.

The excess length came to light in the 1980s, when assisting a PO/BT engineer clear a fault on it.