Delta loop gives me early warning!

The HF bands have been quite lively today, so I tried a couple of CQs at 28MHz using my 20m vertical delta loop on its first harmonic.  This normally yields a low SWR of about 1.2:1.

But on sending the FT8, I could see the SWR was fluctuating, indicating an intermittent break in the wire.

At 14MHz, because the current reaches zero at the apex, where there is a lot of physical stress due to high winds, a break there has no effect on the normal, efficient operation of the antenna as the pair of closely-spaced verticals that it effectively is, as the model in MMANA-GAL shows:

At 28MHz, the current isn't zero, and is in fact fairly large at the apex, which is why the break, which is almost certainly there, is evident at this frequency but not 14MHz:

So, there's a little job for me over the weekend.  Luckily, it's a very easy and quick job.

The break was indeed at the apex. 

Job done in about 45 minutes.  Total wire length = 22.0m.  Length from feedpoint to apex = 5.43m (feed to apex measured before strain relief loops are formed, which results in the length becoming a little shorter later.  Similarly, total wire length is that from which the strain relief loops are later made, i.e. no need to add a bit extra for strain relief loops).  SWR at 14MHz and 28MHz is very good, and low across both bands:

Strain relief at apex now uses the antenna wire looped over a simple insulator, then taped with self-amalgamating tape, and a single cable tie to keep it together.  This affords slightly less stress on the wire, but the previous, cable-tie and loop only arrangement lasted in fierce winds for over 4 years anyway.