Antenna Problems with K3 firmware 5.57

[Bill Frantz]

I have a 160M full length dipole. The center used to be at about
40 feet. About the time version 5.57 of the firmware came out,
the mast supporting the center feed point broke and I re-rigged
with the center at about 22 feet. (I hate debugging when there
were two changes.)

Measurement of the lowered antenna with a MFJ-259B antenna
analyzer show 1.2:1 SWR match at 1.87 KHz with 2,5:1 and 3.3:1
at the band edges from the shack. Because of the mismatch at the
band ends, I usually use the antenna using the KAT3A tuner
installed in my K3. (I don't have similar figures for the higher
antenna, but the SWR curve matches the computer model.)

I used the antenna with modest success including 10 DXCCs and 41
US states confirmed.

When I installed 5.57, I noticed significant power drops on
160M, stepping down from 100W to 25W and 50W. With the help of
Elecraft support, we have concluded that the antenna is
irritating the new firmware feature:

     K3 MCU 5.57 / DSP 2.88 / FPF 1.26, 2-16-2017

     * FASTER RESPONSE TO HIGH REFLECTED POWER. The transceiver
will now reduce
     drive more quickly when transient conditions cause
excessive reflected
     power. This roll-back response does not change the set
power level (PWR
     knob); instead, it reduces the drive to get below the
reflected power
     limit. This means that once a better match is restored
(such as when using
     an antenna tuner), the set power level will automatically
be restored.

When I backed off to version 5.50, 160M performance seems to
have gone back to what it was.

Measuring the open-circuit voltage on the antenna feed line with
a oscilloscope shows about 600mV peak to peak. I estimate, based
on eyeballing the trace and measuring other antennas, that about
200mV of that is 60 cycle hum. Elecraft support thinks this is
enough voltage to affect the internal K3 SWR bridge.

The voltage is most likely from AM broadcast stations. The
strongest station, measured with the P3, is KLIV (5000W) about 8
miles from my QTH.


So now, the question is what to do about the problem. Several
things I have been able to think of are:

* Use a broadcast band filter. DXEngineering has one that will
handle 200W, but it wants 50 ohms impedance on both ends.
Therefore it is recommended to install it between the power amp
(internal to the K3) and the tuner (also internal to the K3). I
might be tempted to insert one in the K3, but the tuner and
power amp are connected through a header, not a simple cable I
could reroute to install the filter.

* Get a KAT500 and the filter. Expensive, but straight forward.

* Sometimes the power remains at 100W when the internal tuner is
in bypass mode, but not always. It might remain stable enough
for QSO, but flipping back and fourth seems to be a pain.

Does anyone have any suggestions?

Thanks - Bill AE6JV

-----------------------------------------------------------------------
Bill Frantz        | Concurrency is hard. 12 out  | Periwinkle
[phone]      | 10 programmers get it wrong. | 16345
Englewood Ave
www.pwpconsult.com |                - Jeff Frantz | Los Gatos,
CA 95032

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

[Don Wilhelm]

Bill,

I would try an L/C series trap at the frequency of your offending BC
band station across your antenna feedline.  That should not be dependent
on the IN/OUT impedance requirement of the DX Engineering BC band filter.

73,
Don W3FPR

On 3/31/2017 5:36 PM, Bill Frantz wrote:

>
> The voltage is most likely from AM broadcast stations. The strongest
> station, measured with the P3, is KLIV (5000W) about 8 miles from my QTH.
>
>
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

[Jim Brown-10]

I agree, and you'll need pretty high Q components. Lots of good
capacitor choices at HSC. Another suggestion though. KLIV is omni during
the day, but strongly directional at night. This is quite common on the
AM band. I suggest that you make your strength measurements at night.
Depending on where you are, a different station may be stronger at night
when you're working 160.  Here's a link to the FCC website where you can
look up technical details of any broadcast station.

  https://www.fcc.gov/media/radio/am-query

And here's the listing for KLIV Notice the links near the bottom of each
listing showing pattern and coverage area.  The wild distortions in the
coverage area are the result of varying ground conductivity -- the big
thumb over the Bay is because it's sea water.

https://transition.fcc.gov/fcc-bin/amq?list=0&facid=19531

Another thought, Bill.  A horizontal 160M antenna at 20 ft is a real
worm-warmer, and while 40 ft is better, it's still very low. If you can
do some sort of counterpoise, a top loaded vertical, even if it's short,
could work a lot better.

Here are slides for the talk I've done several times about getting on
160M. It's mostly about antennas and counterpoises. You might have heard
it. :)  The K2AV folded counterpoise seems to work pretty well, and
doesn't take a lot of space.

73, Jim K9YC

On Fri,3/31/2017 2:50 PM, Walter Underwood wrote:
> Use a series-tuned “wave trap” on KLIV’s frequency across your feedline.


______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

[Elecraft mailing list]

This may not be directly related, but I had power indication problems in
that the SWR and FWD power readings on my K3 had become erratic with surging
of indicated power on CW from a setting of 50W up to 100W and over, yet a
constant tune signal was rock steady at 50W. I thought it was an
intermittent antenna problem, then switched to it being an SWR measurement
problem.  I tried everything including two updates of firmware which had
temporary benefit but didn't last. Wrote to Elecraft and carried out all
checks they recommended except (to my embarrassment) for one which I thought
unlikely.  There is a lesson here!!  Yes, it did turn out to be the real
cause and that was to check and tighten the three screws on the bottom cover
for the LPA transistors (p63 in assembly manual).

 

My assumption now is therefore that loose mountings / grounding may have
been causing erratic reactive components in the LPA power circuitry leading
to sprogs which trigger the much improved VSWR sense circuitry to hunt but
never actually find a stable point except when on steady tune up carrier.
Incidentally it didn't show up when running low power (10W) using the LPA
only. It is so easy to jump to the wrong conclusions; in this case where
improvements in the SWR sense loop performance was masking the growing
problem with poor ground of key LPA components. After re-reading the
Elecraft advice again I did actually check the screws (they were very loose)
and tightened them up it all went back to normal.  Disconnecting the cables
from the K3 to connect to the PC for the firmware uploads must have
temporarily disturbed the seating of the LPA screws giving a false and
temporary indication of a fix.  A regular ragchew contact reported much
cleaner rf signal as seen on a pan display after the repair.

 

Dick G3WRJ

 

 

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

[Bill Frantz]

On 4/1/17 at 6:44 AM, [hidden email] (Dick Bacon via
Elecraft) wrote:

>... that was to check and tighten the three screws on the bottom cover
>for the LPA transistors (p63 in assembly manual).

I got the same advice, but mine were quite tight. If you have
problems with erratic SWR or Power readings, check those screws
before sending an email to support. :-)

An update on the broadcast station interference problem aka the
antenna problem.

I looked at KLIV's data in the FCC database (thanks Jim K9YC for
the links). It turns out that after sundown, KLIV changes the
pattern of its antennas to put most of the power directly toward
my QTH.
<https://transition.fcc.gov/Bureaus/MB/Databases/AM_DA_patterns/429300-67110.pdf>.
No wonder I have a problem.

Jack, W6FB wondered how long a coax stub for 1590 KHz would be.
A quick back of the envelope calculation came up with a bit more
than 100 feet. I pulled out a piece of RG-58 labeled 104' and
hooked it up as a stub. The problem get a lot better, but it had
a noticeable effect on the bottom end of the 160M band. The stub
is far from being a notch filter.

Adding another piece of coax got it below the 160M band, but I
worry about suppression of odd harmonics at higher frequencies.
I think I have used the antenna for QSOs on every amateur band
from 160M to 6M inclusive. So on to something that took longer
than 10 minutes to try.

I dug through the junk box and found a 560 pF 1KV cap and a
T68-2 core. Quality time with the ARRL Handbook said to use 56
turns on the core, which I did with some #30 wire. So far, knock
on wood, this seems to get proper power output on the entire
160M band. And the core only gets warm if you hold the key down
for a long time. I'll need thicker wire before the next 160M
RTTY contest.

73 Bill AE6JV

-----------------------------------------------------------------------
Bill Frantz        | Security is like Government  | Periwinkle
[phone]      | services. The market doesn't | 16345
Englewood Ave
www.pwpconsult.com | want to pay for them.        | Los Gatos,
CA 95032

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]