Hi Stefan,

> That is very strange! Thermal capacity and time constants should be the 
> same when heating up or down.

Well... yes, that seems to make sense. :-)

> How much antenna current do you get at 100 W RF power?

It depends on the time of year and temperature, since the loss
resistance is dependent on those factors. On the last day of testing
it should have been about 1.1A.

I do not normally keep an RF ammeter in the antenna lead, but I
think I need to put one out there again to check whether antenna
current increases when I see current on the feed line increasing.

> If you say you have a scopematch, i assume you mean that you are the 
> curves of voltage and current on an oscilloscope, right?

Yes, I always watch the curves of voltage and current on the feed
line (measured at the transmitter end of the line).

> I have thoughts 
> about EMI into the power supply or exciter but this is unlikely since 
> you have a decoupling xfmr at the antenna feed point.

It's quite possible, but how could EMI into the power supply or
exciter result in feed line current increasing while at the same
time voltage is decreasing? I could understand a power change, but
this isn't that, it seems to be a change in antenna / ground system
resistance.

> Is your antenna earth separated from the mains earth? If not, try to 
> separate it and see if the effect is still there, apart from the lower 
> antenna current.

Actually it is not. I do have the isolating transformer at the
antenna base, but I do not think that accomplishes much at my station!

The 'T' antenna for LF is suspended between two metal towers. Each
tower is grounded, and around one of the towers is a quite extensive
system of wire radials because that tower is shunt fed on 1.8 MHz.
The lightning ground for both towers plus the radial system are all
tied together and this is the ground for the LF antenna.

The trouble is coax cables going to other antennas on both towers
are connected to equipment in the shack, which of course is
connected to the mains earth. So we cannot say the LF antenna earth
and mains earth are separated. In fact, we can say it may be more
like a big loop!!!

I can try to disconnect all of those cables temporarily but I am
quite certain this will shift resonance of the LF antenna. It will
take some time to get it resonated for another test. It is quite a
lot of work to change taps on the coil, as I must completely
disconnect it and remove it from the large plastic drum to be able
to reach the taps. Not the ideal setup at all, but it is what I
could do.

I wonder if I should try to temporarily connect the coax outer to
the antenna ground at the base of the LF antenna? At least that
would provide a direct return path.

> And maybe it is worth to control the DC supply voltage of the exciter 
> and PA during the effect starts.
> If the PA runs at 12V, use a battery and a floating signal generator.
> These are my next thoughts.

This happens with two very different exciters. One is the U3S. The
other is a HF transceiver through a home made down converter. The
affect I am seeing is the same using either exciter.

Make that three exciters. Before it died, I could use my HP 3325B as
an exciter. I saw exactly the same thing when testing with that.

I can easily control voltage to the PA but not so easily to any exciter.

The PA will run at 12V but does not produce enough output at that
voltage to see the problem. Below a certain RF power threshold there
is no problem!

73,
Paul