[Elecraft] KPA500 RFI problem

[Guy Olinger K2AV]

Part of the difficulty is an assumption that 1) there really is an antenna
that will work on all bands, 2) that balanced feedlines do not have common
mode current, and 3) that just about any feedline and transformer
combination will work.  There is an awful lot of "all-band" stuff that is
really snake oil, and depends on LOSS to push down the SWR.  True,
efficient all-banding is work, requires the good components and scanning RF
analyzers to work and adjust them into shape.

Both a transformer or any wire plus ferrite device and an auto-tuner that
are subjected to the stresses from an "all-band" application can 1) be
ineffective due to no in-range solution at a given frequency, 2) overheated
due to common mode current or high current, or 3) in the extreme destroyed
by heat or arcing paths.

An antenna that is tunable on all bands including WARC has an element of
luck to it, even with antennas claimed to be all-band.  To be efficient as
well means that extremes in R, X and Z are presented in full strength to
stress points, and are not soaked up by potentially lossy elements in the
antenna system such as feedlines, transformers, and wire/ferrite devices.
Destructive possibilities are made worse by cores and coax that barely can
stand constant key down 100 watts with extreme R, X, and Z.  While I doubt
this for the KAT500, an awful lot of stuff peddled as "baluns", "ununs",
etc., is really cr*p.  This is particularly true for the inexpensive
devices.   Add to this some unlucky antenna implementations.

Now the odds of working at 100 watts and then failing at 500 watts are
remarkably common.  You have multiplied prior heat loss by 5.  You have
better than doubled current at high current points, and you have better
than doubled the voltage at possible breakdown points.   Stuff that was
made as cheap as possible, and blow-up tested at 100 watts for the low end,
now becomes questionable.

And even if you were previously lucky, adding a half dozen feet of
miscellaneous conductors to insert an amp and/or auto-tuner to the string
can present R, X, or Z that is now out of range or destructive, especially
considering the higher power.

Just wanted you to know that if you are doing one antenna multiband, you
are in the swamp and there are alligators.  Nothing that can't be managed,
you just need to know that all-band antenna design has sharp teeth, and
will bite.  Forewarned and all that.

There are other analyzers that can do this, I have an AIM4170 scanning RF
analyzer.  The 4170 can be set to scan from 3 to 30, or 1.5 to 30, or any
continuous range of frequencies that is your target range.

The entire range at one time can be displayed graphically on the PC
connected to the 4170.  Ham bands can have a darker gray background to make
it easier to see it all at one time.  Then the sweep display of R, X and Z
will tell you what is going to happen.  One can dink around with wire
lengths and open wire lengths, and inclusions of 4:1 transformers (or not).
 What you are looking for is neither really high or really low extremes of
 R, or X or Z in the ham bands.  High extremes will defeat tuner matching,
cause arcing in connectors, and low extremes will cause overheating when
presented to transformers.

You need to do the viewing anywhere you intend to insert a device, and at
either end of coax runs.  In particular, before coax do the all frequency
scan WITH and WITHOUT a 4:1 or 9:1.  Also consider running dual coax, with
the two center conductors as a balanced line with the shields connected
together at both ends and grounded at entry to the shack.  Dual RG-11 (150
ohms balanced) is good for terminating open wire for coax runs, and may
moderate impedances (and loss) seen at the tuner end.

When an extreme is close or in a ham band, seemingly small changes in
feedline or even antenna height and changes in proximity to conductors,
just a small change can move an extreme point  just enough to hit a
frequency you are using.

If you haven't bought a 4170 or equivalent, or borrowed a friend with a
4170 or its equivalent, basically you are flying blind, and just hoping you
didn't drop an extreme in a band you want.   An SWR meter is really not
sufficient to the problem unless you are lucky.

73, Guy
K2AV

On Tue, Jan 8, 2013 at 8:06 PM, Ron W3ZV <w3zv at verizon.net> wrote:

> At least I'm pretty sure its an RFI problem. I have a K3, KAT500 and the
> KPA500 connected as recommended with Elecraft DB15HD cables. The antenna is
> a 120ft. doublet fed with ladder line to a 4:1 balun and 10ft of coax to
> the KAT500. I have a comment point ground to a ground rod six foot from the
> shack. I have operated the amp at 400+ watts from 80m to 6m. The only band
> that gives me a problem is 12m. When I am operating at 400 watts ssb, the
> amp will fault with a high swr indication within 5 to 20 seconds.
> Immediately prior to the fault the swr on the amp display was 1.2:1.
> Reducing power to below 300 watts seems to avoid the problem.
>
> Assuming the consensus is that this is RFI, what would be the most likely
> suspect? Coax cables in/out? DB15HD cable? Power cable? What's the easiest
> way to diagnose and correct the problem? Thanks
>
> Ron W3ZV
>