[Elecraft] KPA-1500 SWR fault above 700-800 watts on 160m

[Alan - G4GNX]

Ken, sorry this is rather long, but here goes...............

I had a similar problem in the UK, with a OSCFD antenna, purchased from 
a UK company. I didn't question the theory and just assembled it 
according to instructions.

As you would expect the antenna is made from two differing lengths of 
ruggedised, plastic coated copper wire, which is fed by a 4:1 Guanella 
balun, rated at 400W. The balun is potted in a weatherproof box and is 
fed via the customary PL259/SO239 connector with Mil Spec RG58 coaxial 
cable. About 6' down from the feed point is a common mode choke 
(sometimes called a sleeve balun) which consists of 8 ferrite inline 
cores, around the co-ax and held in place with heat shrink sleeving and 
cable ties.

The rig was just the K3S/100W with internal tuner and all was well for 
many months, mostly using 40 metres SSB. A while later, I added the 
KPA500/KAT500 combo which I ran for a couple of very short periods at 
400W to test. For a weekly net, I used only 200W and all seemed OK for a 
while, until suddenly the VSWR started to increase over the period of an 
'over'. Over a few weeks this got more frequent and I would have to 
re-tune part way through an 'over'. At that time, I wan't sure whether 
it was an antenna issue, or the KPA500/KAT500, although I suspected that 
the KPA500 was getting too hot, so I posted a question on a forum and 
was contacted by Jack Brindle of Elecraft, who kindly looked at the 
KPA500 fault reports and sent me a comprehensive explanation of his 
findings based on the figures and his experience, which concluded that 
the antenna was almost certainly the culprit.

I ordered a new 1KW Guanella balun and some much bigger ferrites, with 
the intention of replacing the RG58 with Westflex 103. In the meantime I 
soldiered on with the original setup until one day all hell let loose, 
with the KPA500 and KAT500 fault lights flashing and even the K3S got 
'ticked off'. At this point I switched to my HF vertical, to keep me on 
the air.

When I dismantled the wire antenna, I expected a balun fault, but it was 
actually OK. On further checking, after disconnecting the co-ax from the 
balun, the open ended co-ax showed a DC short on a test meter!

I used the Rig Expert 600 to determine exactly where the short was and 
it turned out to be exactly where the sleeve ferrites had been, BUT 
there was no discoloration of the outer covering, or signs of melting, 
so I knew it was unlikely that the ferrite cores had got heated. I cut 
the section out of the co-ax and started to take it apart. With the 
outer covering all removed, there was still no sign of discoloration of 
the screen braid, but as I tried to remove the braid, it was obvious 
that there had been considerable heating as the center section of the 
braid has melted into the center conductor insulation, and somewhere in 
there, it has made it right through to the center conductor.

Unfortunately I haven't yet found the exact point of the short, because 
I left the partly dismantled piece of inner co-ax on a table and one of 
my cats seems to have run off with it! :-D
When I eventually find it, I'll continue investigating.

In the meantime, one theory is that the common mode choke presented a 
high impedance to RF coming back down the co-ax and this in turn 
produced a very high voltage causing an arc between inner and outer and 
as the characteristics changed, in turn a high current node was created, 
which caused the heating effect. Another theory is that it is just a 
crappy piece of cable and it just failed.

Unfortunately this is not the sort of fault that can be detected in its 
development stage by a low power antenna analyser, it needs a fair 
amount of power.

This long tale may just give you some clues to your possible issue on 
160m although on a different power scale..

73,

Alan. G4GNX


------ Original Message ------
From: "Ken Ramirez de Arellano" <kp3mm.dx at gmail.com>
To: elecraft at mailman.qth.net
Sent: 02/09/2020 22:41:30
Subject: [Elecraft] KPA-1500 SWR fault above 700-800 watts on 160m

>Every bit of antenna hardware ( coax, balun, insulators, etc. have been
>changed out or proven good. Connections are all solid.  My KPA1500 will
>transmit above 800 watts for about 15-30 seconds and then the SWR jumps up
>and it faults out for high VSWR.  I can run 1500 watts on all the other
>“Legacy” bands (10,15,20,40,80) with no problems. I’ve forced an STU
>relearn and still the same issue. The antenna B amp is a Clipperton L and
>that transmits at a KW out and I see no change in VSWR in the external
>wattmeter using all of the same cabling. It’s starting to look like
>something heating up in the KPA1500. Is anyone aware of a known issue for
>this with the KPA1500 or a known fix? I would rather do a field repair
>instead of sending to Elecraft from KP4.  TIA Ken, KP3MM
>--
>Ken Ramirez
>KP3MM
>