Initially I tried TNP's method of a few turns of thick wire (actually
about 20 turns and 1mm^2) on a small plastic bobbin (a wheel
'borrowed' from one of the trays in the dishwasher!), connected it via
a 'Quicktest' connector (see http://tinyurl.com/qffkxxs ) to a cooker
type panel supplying a 7.5 kW pottery kiln. Banged down the lid, quick
flash from the neon, blew the 13A fuse, tripped the 50A MCB and the
RCD in the dedicated CU, recovered the little magnet, but zilch! Not a
Gauss (or should that be an Oersted), well, nothing significant,
On to plan F (or is it plan G now?) Whichever, it was a combination of
Bob Minchin's and Martin Brown's suggestions of getting a much smaller
Nd magnet and/or chipping a bit off one that I'd got. The Nd magnets I
had were 4mm dia. and 6mm long, and I'd bought a pack of 10 from
First4Magnets so plenty to play with. I chopped one into three
more-or-less equal lengths by holding it in the wire-cutting section
of a pair of pliers at the point I wanted it to break (2mm) and then
squeezed the pliers in a small bench vice. They snapped/cleaved
perfectly in just the right place, pretty much straight across. So I
now had three Nd magnets, ~2mm long. I tried two together, but they
were still too strong, but one on its own, with a couple of short bits
of mild steel rod to provide the extra length and weight, and Bingo!
it works perfectly, and the gauge calibration is still OK.
I'm delighted. Many thanks to everyone for their helpful suggestions.