Elecraft

KPA500 RFI Problem

Classic

List

Threaded

8 messages Options

w8fn

KPA500 RFI Problem

Howdy, all. I'm attempting to cure a weird RFI problem that is affecting
a K3/KPA500 combination. I've done some work to troubleshoot the
problem, but before I get too much deeper into detective mode, I'm
looking to see if anyone else may have experienced this problem and
maybe even cured it.

It's obviously a common mode RF problem. It appears only on 20 meters
when I point the antenna in one particular direction. With the help of a
MFJ RF current probe I've determined that for some unknown reason the B
port control cable on my microHAM Double Ten switch is picking up lots
of common mode RF. I've put a bunch of ferrites on the cable and this
has reduced the current on the cable but not eliminated the RFI problem.
It doesn't seem to be coupling from one of the other antennas, as I've
disconnected them all and the problem persists. Also, the RFI is at its
worst when the antenna is pointed at right angles to all the other
antennas. I'm stuck with the physical environment and antenna geometries
I have now, so I'd prefer to just find the point of RF ingress that
causes the problem and attack it there.

The problem shows up as a creep in output power when the transmitter is
keyed. At antenna headings away from the trouble zone it takes about 31W
to produce 500W output from the amp. In the trouble zone at key down the
output power quickly climbs to 600W+ and I have to back down the drive
power from the K3 to keep the amp from faulting. I can't find a stable
drive power that will maintain rated output -- I have to drop the drive
power to about 15 or 16 Watts to hold the output to the 300W range
before the output stabilizes.

Accompanying this effect is a modulation of the signal that is reported
to give it a "buzzy" sound, and I also get a pronounced buzz in the
headphone monitor audio from my MK2R+ SO2R box under these conditions.
Even with the drive down to where the amp output is about 100W the
signal still has noticeable modulation on it. This leads me to believe
the problem is in the KPA500. My first guess was that the RFI is getting
in via the serial bus in the ACC cable. I have tried winding varying
numbers of turns of the ACC cable around a 2.4" Mix 31 ferrite core with
no noticeable result.

Does anyone have an idea of where the RF could be getting in to cause
this kind of problem? I'd greatly appreciate any reports from those who
may have encountered this situation and would especially welcome
suggestions on how to find the cause and fix it. In the interest of
brevity, I've probably left out many pertinent details that may give
further clues, but for now let's see what we can do with the information
I've given so far.

Thanks & 73...
Randy, W8FN
______________________________________________________________
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

Jim Brown-10

Re: KPA500 RFI Problem

On 1/22/2014 7:36 PM, Randy Farmer wrote:
> I've put a bunch of ferrites on the cable and this has reduced the
> current on the cable but not eliminated the RFI problem.
"A bunch of ferrites on a cable" won't do much at HF. You must wind
multiple turns through the core to make a dent in RFI at 14 MHz.

I think it's unlikely that the problem is generated in the KPA. I
suspect either the K3 or the Microham box, so I would work on choking
the cables connected to those boxes. Also, try disconnecting everything
from this system except the bare minimum -- the mic, the K3, and the
KPA. remove the AUX cable, use the RCA cable from the K3 to the KPA.
{The KPA works fine without the AUX cable -- it samples the first bit of
RF drive it sees, counts the frequency, and switches bands very
quickly.) If that's clean, go back to the AUX cable. If that's clean,
add the Microham.

BTW -- 5-6 turns is about right for 20M for a 2.4-in #31 core.

That 300W number is important information -- it tells us that we only
need about 3dB of suppression to solve our problem. That's not a lot
once you figure out which cable(s) is (are) letting the RF in. .

Another thing to look at is the coax between the K3 and KPA. Is this
first quality coax with a real manufacturer's brand name (Belden, Times,
Commscope) with Amphenol connectors, properly installed, or is it some
off-brand advertised for cheap in QST or CQ, or even from a flea
market? Flaky coax or a lousy connector or a connector poorly installed
can cause things like this.

Also look at the interconnecting audio cables. Many (most?) stereo
cables sold by the big box stores have flimsy shields (or no shield at
all), even when they're charging a lot of money for them.

73, Jim K9YC
______________________________________________________________
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

w8fn

Re: KPA500 RFI Problem

In reply to this post by w8fn

Thanks to all the folks who replied with advice, both on and off the
list. I spent the better part of the weekend working the problem and
thought I'd update y'all on what I found.

It turns out there were two major problem areas. The first was RF
getting into the K3 via the ACC jack from the connections between the
radio and the KPA500, the XV144 and the microHAM MK2R+ box. The KPA500
is NOT affected by the common mode RF on the ACC cable. The power
gyrations were clearly due to RFI effects within the K3. I *mostly*
fixed this by building a short extension cable for the ACC jack and
wrapping six turns of it around a 2.4" mix 31 toroid. This lets me use
up to 500W output at the problem antenna heading without making the
radio go nuts. If I increase power slightly it still begins to act up,
so I haven't completely cured the problem. It's easier to just back the
power down a few watts than it is to figure out how to get another few
dB suppression on this path. I plan to live with it.I still have a
little RF getting into the receive audio line at the worst case antenna
heading and causing a bit of hum on the audio, but this is not really a
big deal.

The second problem turned out to be the 13.8V power supply I use for the
antenna switching relays and powering the Station Masters. This thing is
a cheap knockoff of an Astron supply that uses a 723 voltage regulator.
RF was apparently getting into the supply and modulating the DC output.
This caused the antenna relays in the Station Master to buzz. This was
undoubtedly the source of the "buzz" on my signal reported by listeners.
I took several turns of the output DC cable around another mix 31 toroid
near the supply output terminals and this seems to have cured the problem.

Thanks again to those who weighed in with suggestions. You gave me a lot
of troubleshooting approaches that let me home in on the problem pretty
quickly once I got started.

73...
Randy, W8FN
______________________________________________________________
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

Don Wilhelm-4

Re: KPA500 RFI Problem

Randy,

You might have better success with applying the 2.4 inch mix 31 toroids
to your feedline rather than trying to eliminate RF intrusion on every
cable in the shack.

In other words, kill the common mode current at its source and you may
not have to add additional ferrites on all your shack cables. 3 to 5
turns through a stack of 5 2.4 inch mix 31 cores at the point where the
feedlines enter the dwelling may do the job nicely. Another similar
choke at the antenna feedpoint(s) may be helpful as well.

The place to deal with RF-in-the-Shack is at the antenna field, not in
the shack.

Just for the record, OCF antennas are infamous for these problems, and
with any antenna where the feedline are run under the radiator rather
than dropping at 90 degrees from the feedpoint for at least 1/4
wavelength can be expected to pick up radiation from the antenna and
feed it into the shack as common mode currents.

73,
Don W3FPR

On 1/26/2014 6:42 PM, Randy Farmer wrote:

> Thanks to all the folks who replied with advice, both on and off the
> list. I spent the better part of the weekend working the problem and
> thought I'd update y'all on what I found.
>
> It turns out there were two major problem areas. The first was RF
> getting into the K3 via the ACC jack from the connections between the
> radio and the KPA500, the XV144 and the microHAM MK2R+ box. The KPA500
> is NOT affected by the common mode RF on the ACC cable. The power
> gyrations were clearly due to RFI effects within the K3. I *mostly*
> fixed this by building a short extension cable for the ACC jack and
> wrapping six turns of it around a 2.4" mix 31 toroid. This lets me use
> up to 500W output at the problem antenna heading without making the
> radio go nuts. If I increase power slightly it still begins to act up,
> so I haven't completely cured the problem. It's easier to just back
> the power down a few watts than it is to figure out how to get another
> few dB suppression on this path. I plan to live with it.I still have a
> little RF getting into the receive audio line at the worst case
> antenna heading and causing a bit of hum on the audio, but this is not
> really a big deal.
>
> The second problem turned out to be the 13.8V power supply I use for
> the antenna switching relays and powering the Station Masters. This
> thing is a cheap knockoff of an Astron supply that uses a 723 voltage
> regulator. RF was apparently getting into the supply and modulating
> the DC output. This caused the antenna relays in the Station Master to
> buzz. This was undoubtedly the source of the "buzz" on my signal
> reported by listeners. I took several turns of the output DC cable
> around another mix 31 toroid near the supply output terminals and this
> seems to have cured the problem.
>
> Thanks again to those who weighed in with suggestions. You gave me a
> lot of troubleshooting approaches that let me home in on the problem
> pretty quickly once I got started.
>

______________________________________________________________
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

w8fn

Re: KPA500 RFI Problem

I'm stuck on a very small lot with a pretty ugly RF environment. My
station is on the second floor of the house and the SteppIR is mounted
on a 10' roof tower directly over the station location. The feedlines
and control cables go down the roof and then enter under the eaves from
near ground level and are routed back up to the attic on the wall
immediately behind the radio station, where the antenna switching and
control cable interfaces are located.

I do have good quality chokes (Balun Designs) installed at the antenna
and also at each of the two outputs of my Double Ten switch that feed
the radios. For some weird reason I'm getting strong coupling only on 20
meters and only when the antenna is turned to around 120° to 150°
(exactly normal to the long axis of the house). None of the other wire
antennas or the SteppIR at other headings gives me any noticeable common
mode RF problems.

Unfortunately, in my case the shack IS part of the antenna field and
there's nothing I can do about it short of removing the antenna. When I
retire in a couple of years I fully intend to move to a location where I
can have the antennas properly separated from the living quarters and
radio shack. Until that happens I'll just have to deal with the RF as
best I can.

73...
Randy, W8FN

On 1/26/2014 6:38 PM, Don Wilhelm wrote:

> Randy,
>
> You might have better success with applying the 2.4 inch mix 31
> toroids to your feedline rather than trying to eliminate RF intrusion
> on every cable in the shack.
>
> In other words, kill the common mode current at its source and you may
> not have to add additional ferrites on all your shack cables. 3 to 5
> turns through a stack of 5 2.4 inch mix 31 cores at the point where
> the feedlines enter the dwelling may do the job nicely. Another
> similar choke at the antenna feedpoint(s) may be helpful as well.
>
> The place to deal with RF-in-the-Shack is at the antenna field, not in
> the shack.
>
> Just for the record, OCF antennas are infamous for these problems, and
> with any antenna where the feedline are run under the radiator rather
> than dropping at 90 degrees from the feedpoint for at least 1/4
> wavelength can be expected to pick up radiation from the antenna and
> feed it into the shack as common mode currents.
>
> 73,
> Don W3FPR

Vic Rosenthal

Re: KPA500 RFI Problem

Jim Brown-10

Re: KPA500 RFI Problem

In reply to this post by w8fn

On 1/26/2014 7:18 PM, Randy Farmer wrote:
> Unfortunately, in my case the shack IS part of the antenna field and
> there's nothing I can do about it short of removing the antenna.

Yes, when the antennas are in close proximity, they radiate RF into the
shack, independent of common mode on the feedline. That said, OCF
antennas are a really bad idea in today's world, both because of noise
coupling and the high common mode stress they put on a choke. I don't
know anything about the Balun Designs productcs, but even the best
chokes can be fried by running high power through them to a badly
unbalanced antenna.

Some other thoughts. It is critically important that the shield of all
cables be connected to the shielding enclosure at both ends, and ONLY to
the shielding enclosure. You should make sure that the cables you are
using do that, and it may not be easy. I terminate the signal returns
of all lines in my DB-connectors to the DB connector shell. This
requires that the connector shell be bonded to the shielding enclosure,
and with most gear, that works. It did NOT work with a cable I made to
go between the K3 and a KAT500, so I suspect that there is a Pin One
Problem at the AUX jack one or both of those products. What you are
describing is symptomatic of a Pin One Problem.

Pin One Problems are tough to diagnose, because the places at which they
inject RF are at the whim of the circuit layout designer. The choke
you've put on the AUX cable sounds about right for 20M, and should be
effective from 30-10M. At this point, I'd look at what other cables are
connected to the radio and choke them.

There's another thing that might be causing your problem that I ran into
only a few days ago helping W6OAT solve a similar problem -- his big
SteppIR is about 60 ft above the shack on a tower, and he gets RF
feedback into the radio that locks up CW keying . He's using a Microham
box (don't know model number) and the interface cable between that box
and the K3 is a train wreck waiting to happen. There's a big multi-pin
plugged into the Microham with a lot of wires coming out of it that have
labels on them like Line In, Line Out, Key, etc. but they are all
unshielded, and there's no signal return -- the setup is depending on
the coax shield for that. That works for DC, and maybe even for audio
(if there's no big power transformer nearby), but NOT for RF.

The problem with using coax (or even a chassis bond between the
equipment) as the signal return is that total circuit forms a big loop
consisting of the signal wires, the chassis, and the coax shield. You
are certainly in the near field of your antennas, and in the near field
of a current maximum (in this case, the antenna feedpoint), the magnetic
field dominates and couples into that open loop, and the strength of the
coupling is proportional to the loop area! In addition, that loop forms
a receiving antenna, which picks up the electromagnetic field.

I have no idea where Rusty's cable came from, but it's clearly
commercially made, and seems to be specific to the Microham and the K3.
Cable adapters like this may be simple to hook up, but they are a
sitting duck for RFI.

73, Jim K9YC
______________________________________________________________
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

w8fn

Re: KPA500 RFI Problem

Thanks much for your thoughtful and informative reply, Jim. My comments
are interspersed.
On 1/26/2014 10:32 PM, Jim Brown wrote:

> Some other thoughts. It is critically important that the shield of all
> cables be connected to the shielding enclosure at both ends, and ONLY
> to the shielding enclosure. You should make sure that the cables you
> are using do that, and it may not be easy. I terminate the signal
> returns of all lines in my DB-connectors to the DB connector shell.
> This requires that the connector shell be bonded to the shielding
> enclosure, and with most gear, that works. It did NOT work with a
> cable I made to go between the K3 and a KAT500, so I suspect that
> there is a Pin One Problem at the AUX jack one or both of those
> products. What you are describing is symptomatic of a Pin One Problem.

I agree. It seems the K3 is responding badly to the RF on the ACC cable,
whatever the coupling mechanism. I can't imagine RF getting into any
other component of the system that would cause the K3 to change its
power output. If the KPA500 was being bombed it would probably show some
truly bizarre behavior. It seems to just faithfully amplify whatever RF
gets delivered by the K3. I haven't looked real closely at the K3, but
since they're using D-sub connectors connected to printed circuit boards
a Pin One problem is practically guaranteed. I'd sure like to see
manufacturers return to using chassis mounted connectors with internal
connections to the PCB via multiconductor cables that plug into headers
on the board. I don't think it will ever happen, though.
>
> Pin One Problems are tough to diagnose, because the places at which
> they inject RF are at the whim of the circuit layout designer. The
> choke you've put on the AUX cable sounds about right for 20M, and
> should be effective from 30-10M. At this point, I'd look at what other
> cables are connected to the radio and choke them.
I'm using the microHAM factory cable. See below.

>
> There's another thing that might be causing your problem that I ran
> into only a few days ago helping W6OAT solve a similar problem -- his
> big SteppIR is about 60 ft above the shack on a tower, and he gets RF
> feedback into the radio that locks up CW keying . He's using a
> Microham box (don't know model number) and the interface cable between
> that box and the K3 is a train wreck waiting to happen. There's a big
> multi-pin plugged into the Microham with a lot of wires coming out of
> it that have labels on them like Line In, Line Out, Key, etc. but they
> are all unshielded, and there's no signal return -- the setup is
> depending on the coax shield for that. That works for DC, and maybe
> even for audio (if there's no big power transformer nearby), but NOT
> for RF.

According to the microHAM drawing, the factory-supplied K3 interface
cable (type DB37-EL-K3R) is composed of individually shielded lines that
have their shields tied back to the ground side of the RCA DC power
pigtail (which I am NOT using) and are also connected to the DB37
connector shell and the pin 15 Ground line. The Line In, Line Out and
Mic cables run their returns to dedicated pins on the DB37 connector. I
opened my MK2R+ up to have a look and of course found the usual PC board
mounted D-sub connector stacks with no direct connection to the metal
enclosure, although they do show DC continuity to the case. I put a
small solder lug under the jackscrew sockets at each of the D-sub
connectors and ran short #16 bus wires between all the connectors on the
outside of the case to bond them together and then took a ground wire
from the nearest connector frame to a heavy copper bus bar that connects
the ground lines from radios, amplifier, and other major boxes in the
station. I don't *think* RF is getting into the MK2R+ in my case.

>
> The problem with using coax (or even a chassis bond between the
> equipment) as the signal return is that total circuit forms a big loop
> consisting of the signal wires, the chassis, and the coax shield. You
> are certainly in the near field of your antennas, and in the near
> field of a current maximum (in this case, the antenna feedpoint), the
> magnetic field dominates and couples into that open loop, and the
> strength of the coupling is proportional to the loop area! In
> addition, that loop forms a receiving antenna, which picks up the
> electromagnetic field.

How well I know. I spent five years doing CDMA base station design for
Motorola, and we were constantly fighting unintended radiation problems
that were caused by improper digital cable routing and termination.
>
> I have no idea where Rusty's cable came from, but it's clearly
> commercially made, and seems to be specific to the Microham and the
> K3. Cable adapters like this may be simple to hook up, but they are a
> sitting duck for RFI.
I would think it's the microHAM factory cable I've described above.
Thanks again for the help.

73...
Randy, W8FN
______________________________________________________________
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