[Elecraft] KPA500 faulting on high VSWR on power rise

[Rick WA6NHC]

One exception is when the dipole is used on multiple bands, then a 4:1 
may be required.  That was the case for my 370' 80M EDZ (two opposing 
5/8 wave, center fed).  That is specifically a non-resonant antenna on 
any ham band but it performed best with the 4:1 instead of the 1:1.  It 
was a compromise either way, this was the better choice since I needed 
that antenna to work on all bands.  If it was only 80M, the 1:1 was the 
better choice.

While I am not a fan of MFJ so consider it suspect in this case, I also 
believe that since the common mode choke (referred to as a balun in this 
thread) should GREATLY exceed the voltage expectations for simple 
overhead.  The cost differences for higher ratings are not significant.  
In my stations, I use the 10KW rated devices from DX Engineering (KPA500 
amp) and I've had no issues with them, even at extreme SWR mismatch (QRO 
on 160M into the above dipole, 20:1 unmatched with a Heathkit tuner).  
My new station will be entirely resonant, so a 1:1 CMC will suffice but 
I'll still use the high power devices.

The bottom line here is that the antenna in this thread is not resonant 
(because of the feed choice, which is understandable for multiband use), 
the voltages presented will be extreme, component overkill (and a more 
qualified tuner) should be used.  It's not a fault of the amp, but of 
the matching network.

It is also KEY that the LEAST amount of coax is used (under 10' to the 
final) because the high SWR is present there as well, so the losses will 
be extreme and it could even exceed the voltage limits of the coax, 
depending on brand, age and type.  Been there, done that (darned HOA 
limits).  This also means that the coax should be tested, from time to 
time, until proven worthy.

Rick NHC


On 2/24/2018 9:21 AM, William Shanney wrote:
> I have found that a 4:1 balun is seldom the proper choice for a ladder 
> line interface. I measure the impedance at the interface and then 
> select the balun type (usually a 1:1). The biggest problem I've had is 
> the 4:1 transforming the impedance too low for the tuner. This can 
> result in over stress on the components as the power is increased. If 
> I don't have an easy to match ladder line interface I add or subtract 
> line length until I get one.