Some of you will have noticed that I have been venting my frustrations about
the title subject on the reflector recently and that I mentioned an
interesting anomaly in the earthing arrangement, so over the weekend out
came the deerstalker, cape and meerschaum and the investigation commenced.
The anomaly is this; after finishing reworking a number of things as
detailed in an earlier posting, the there was a sensible increase in antenna
current of about 30 - 40%, so I set about tidying things up round the
'dogbox'. Then to my astonishment the next test produced a further increase,
the resultant  increase totalling about 75% and the BK could be driven right
into saturation. Obviously something had changed, but I had not consciously
altered any of the settings.
("Curiouser and curiouser!" said Alice.)
A visual inspection of the dogbox in the bosky twilight turned up nothing
obvious until I caught a foot in the earthing cable from the point where the
co-ax was terminated at the coil, and after the usual 'oh bother!' (or words
to that effect) I noticed that the cable did not appear to be attached to
the earth point sticking out of the ground.
Attached it and returned to shack. Current down again!
Huh????? (Followed by lots of very rude words indeed)
Just for the hell of it went out into the gathering dark and broke the
connection.
Current increase back again!
Eureka!
(Greek for "somebody get me a towel")
The first part of the puzzle was, where was the TX finding its earth from?
The TX system is complete in itself,  mounted on a bit of Conti-board with
no mains earth as such (or so I thought)  and was not connected to the
station earth point (which comes off the same system as the external point,
at the apex of a fan of 25sq. mm.earth wires and bonded two ways with
10sq.mm cable to all the plumbing in the house, four 4ft earth rods and a
couple of disused electrical earths.)
Eventually discovered that the TX and coil were being earthed to the mains
(a) through the switched mode PSU suppling the IC706, divider/exciter and
the relays, (b) via the earth on the computer COM port, and (c) via the
co-ax to the receiver - which is connected to the station earth.
The resistance  measured between the station and external earth points was
measured today as 0.011 Ohms by passing 15A at 12v between them and a known
low resistance of 0.1 and measuring the voltage drops.
Earthing the TX to the station earth point as well got another few percent
of antenna current, but earthing the coil directly to the external point
drops the current back to where I started and the BK goes unstable as it
approaches maximum output.
Could this be a variation of the old audio amp problem cured by using one
common earth point? Or - which is much more likely - is there something
obvious which is staring me in the face yet still unseen?
This is definitely a three pint problem, and in the absence of a brother
called Mycroft I am appealing to all you helpful people out there to guide
me towards an explanation.

Ian GI8AYZ/MI0AYZ

P.S. Remarks along the lines of "Leave it alone while it's working" will be
treated with a certain amount of disdain, since I have already come up with
that particular way out!