Owen Duffy wrote:

...
It is often overlooked that the power of an SSB telephony signal
averaged over a longish time (but not accounting for the duty cycle of
"overs") is around -15dB (a little less for speech processing), so the
heating effect of 500W PEP telephony might be more like 15W.

Owen

Hmmm, I just may pull that balun down and run a full KW into it hooked
to a dummy load to look at some temp readings over a range of bands/over
a time span ... maybe the thing has been dumping more heat than I am
aware of, time anyway to check out the system ...

JS