I get your point about my errant use of the term “ground” in this case.
I think what you clarified is what I had intended to say, that a coax with
shield not tied to board “common” on both ends does indeed have a long,
variable return loop and this is very much open to RFI pickup, the same way
“twisted pair” or “balanced parallel lines” become antennas if they are not
properly terminated on both ends (and if currents in both parallel lines are
not equal and 180 degrees out of phase with respect to each other).
Gary
> On Oct 2, 2018, at 5:00 PM, Jim Brown <k9yc@audiosystemsgroup.com> wrote:
>
> On 10/2/2018 12:10 PM, Stuart Rohre wrote:
>> Another idea for the cause of the noise clicks is a bad ground in
>> interconnecting cable for this circuit.
>
> Yes, but please don't call it "ground" -- it's NOT! Rather it's a RETURN for
> the current in the center conductor, making it a transmission line. When it's
> a transmission line, all the return current flows in the shield (or the other
> conductor if it's parallel or twisted pair). If the return is missing, it's
> not a transmission line, so return current flows on whatever path the rest of
> the wiring provides, like the chassis. This causes a large inductive loop to
> be formed by the center conductor of the coax and the large return path, and
> that loop couples the magnetic field from other currents in the radio.
>
>> Try a clip lead end to end of the patch cable.
>
> That's a start, because if it closely follows the path of the coax, it
> reduces the size of the loop. But it's still a loop.
>
>
>> Good Luck! Had a similar noise when a custom builder we were using left off
>> a second (ground) conductor for a cross chassis path for a digital clock
>> signal, TTL btw. We wadded a second grounded condcutor wrapped as twisted
>> pair, and cured our noise.
>
> Yes but it's NOT a ground. You simply formed a transmission line.
>
> BTW -- for those who don't know, twisted pair forms a transmission line, and
> if it's really good twisted pair, it's VERY good at resisting noise and other
> crosstalk. CAT5/6/7 cables are four very good twisted pairs, with the added
> (and very important) technique of twisting each at a different rate.
>
> Elecraft uses a short length of twisted pair to tie the input SO239 connector
> of their 500W amp to the circuit board.
>
> A very important design flaw of TenTec's 229-238 series of antenna tuners is
> that it uses the chassis as a return. Palstar made the same mistake in the
> high power tuner I looked at in their Dayton booth, and when I pointed that
> out to the design engineer in the booth, he didn't understand why that was a
> problem!
>
> Ten Tec did NOT make that mistake in their 425 Titan amp -- coax is used for
> 100% of the input and output RF path.
>
> 73, Jim K9YC
> _______________________________________________
> TenTec mailing list
> TenTec@contesting.com
> http://lists.contesting.com/mailman/listinfo/tentec
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|