"Karl Beckman" wrote in
:

First thing to remember is a really radical thought: The power rating
of a voltage balun has to be decreased by the highest SWR factor
existing on the line. That is, a 2 kW rated (at 1:1) balun is only
good for 1 kW at 2:1 VSWR or 200 watts if your wattmeter shows a 10:1
VSWR looking at the antenna. The reason is that the current (heating)
losses in the windings go up as the square of the maximum current -
remember hearing about "I squared R". A 4:1 VSWR means the current
max value is twice the minimum, therefore you'll be heating up the
core four times as much as if you had a 1:! VSWR on the line. It
doesn't matter whether the line is unbalanced coaxial or balanced open
wire.

My comments are addressed to so called voltage baluns. The above
explanation and my comments are not applicable to current baluns.

Your explanation of your Rule of Thumb is valid if the balun is located
at an impedance minimum for the stated VSWR. The explanation does not
apply at all in the opposite case, and to a lesser extent in all other
cases.

In many cases where the balun overheats, it is operating at an extremely
high current, and I^2R losses are not main contribution to heating.


In your case the ratio of rated power divided by working power is
1.25, so the maximum VSWR tolerable before you exceed the dissipation
rating of the balun is only 1.25. Or working the other direction, to
always stay cool you need to find a balun rated for (1200) x (VSWRmax)
watts.

Which also assumes that the balun was properly rated for the 1200W.


The input reactance of the Windom changes as you change the driving
frequency (and band), therefore the VSWR on the feedline does also.
The balun may stay reasonably cool on a few frequencies, but will be
flaming hot on most others.


The core material may only need to reach temperatures towards 200°C to
compromise the balun, the Curie point of some ferrites in use is lower
than 200°C. Thing is that the ferrite heats very slowly, and a 5 minute
test is not likely to raise the temperature to anywhere near the long
term temperature. That is good for us for many modes as they have high
peak to average ratio, but for the low peak to average modes / high duty
cycle modes, the balun core may increase in temperature significantly for
an hour.

The key thing is that assuming balun efficiency is very high is quite
wrong, and when you start to think about balun efficiencies being below
50% under some extreme conditions, you realise the challenge in 5kW rated
devices that are in small boxes and must not reach 200°C.

Owen