On Mon, 25 Jun 2007 12:50:28 -0400, David Jordan wrote:
>Tests were done with 5 watts into a dummy load.
Are you saying that you have this problem transmitting into a dummy load?
>That Kenwood command
>string, inside the encrypted pipe simple tells the TS-480 to go into "send"
>mode. The failure only occurs while the radio is in CW or AM mode.
Hmmm. Maybe RFI in either the 480 or the box that talks to it. Ferrite choke
there? Maybe even a software problem, and not an RFI problem?
FWIW, there is little virtue in using shielded CAT5, especially if its
buried, and especially at HF. I agree with Martin that current on that
shield could cause problems, but I didnt (and still dont) suspect it
because you said only 30M and 40M are involved. Termination of that shield
is critical, and EARTH is the WRONG place to put it. To be effective, that
shield MUST go to the shielding enclosure of the devices at each end.
Otherwise, theres a big loop between the shield, the earth, and that
shielding enclosure, and that loop acts as a receiving antenna.
I also dont see virtue in the K-Com filters. When youre going this far
with Ethernet, youre stretching things anyway, and unless the filters are
really good (to not disturb the signal circuit), they could be doing more
harm than good. All of the RFI to/from Ethernet that Ive seen on cables is
common mode, and that is VERY easily handled with a ferrite common mode
choke that does not disturb the data stream. I have observed differential
mode trash coming from an Ethernet cable, but only around 2M.
As Martin has observed, that shield can also carry noise current, which can
couple into the devices on either end via their pin 1 problems.
Another thought. Ive experienced lock-ups with my IBM T22 keying my
Elecraft K2-100 at power levels as low as 10 watts, with WriteLog sending CW
on pin 4. I traced the problem to a badly designed serial cable that used
shielded straight wires rather than twisted pairs. I fixed the problem by
building a serial cable using CAT5, with one pair for each signaling
circuit. I fixed the pin 1 problem at each end of the serial connection by
terminating the returns of each pair to the DB9 shell. Once I did that, I
could run legal power into a long wire antenna that ran only 2 feet from
that serial cable!
Subsequently, I did a series of tests with shielded and unshielded CAT5, and
shielded twisted pairs, loading that same long wire on all bands 160-10M.
Shielding was totally un-necessary below 20 MHz, but WAS needed above 20 MHz
at the highest power levels. This reinforces what we in the RFI world have
long understood -- TWISTING is FAR more important than shielding at most
frequencies.
Details in my tutorial (the chapter on solving problems in the shack).
CAT5 has 4 pairs, which handles 4 signaling circuits in this manner. Most
rig control uses only pin 2 and pin 3 to common, and most contest loggers
use pin 4 to send CW and pin 7 for PTT. Thats four circuits.
73,
Jim Brown K9YC
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
|