[Elecraft] elecraft,RJ45 Standard

[Jim Brown]

On Mon,5/9/2016 1:21 PM, dave wrote:
>
> > You're missing the fact that although these conductors INTENTIONALLY
> > carry DC and low rate signals, they also can PICK UP RF and AF noise.
> > Twisted pair, when used to carry a signal, is at least as powerful at
> > rejecting RF and AF noise pickup as coax, but ONLY if the pair is
> > dedicated to a circuit.
>
> And also only if the twisted pair is properly terminated in a balanced 
> termination - on both ends. I was field engineer for Bell. I was 
> sometimes totally amazed at just how good twisted pair are at 
> rejecting noise. As good as coax, if not better.

Not true. Yes, cancellation can be improved by balanced termination, but 
ONLY if balanced broadband -- that is, DC to the highest frequencies of 
the interference.  But twisted pair is VERY effective at rejecting 
magnetic coupling even when the interface is unbalanced.

I learned this from solving a severe RFI problem problem with the serial 
interface to my K2, way back in 2003. I was using the "official"Elecraft 
cable, which was parallel conductors, NOT twisted, and my TX antenna was 
a top-loaded end fed wire with a tuner in the shack for 160 and 80M. At 
only 12W, coupling to ,my computer via that serial cable caused it to 
lock up.

I replaced that parallel wire cable with CAT5, using one pair per 
circuit, and making the return connection to the DB9 shells, NOT to the 
"ground" pin (to solve Pin One Problems). Once I had changed the cable, 
I could run my Ten Tec Titan to legal limit with no interference up to 
17 MHz when I intentionally loaded that same antenna on all the HF 
bands. If I used shielded twisted pair, I could run full power up 
through 10M.

>
> But . . . the telco pairs are terminated in carefully balanced 
> terminations. I don't think the typical ham/audio install is so 
> carefully designed. They may be, I don't know for sure. I suspect the 
> terminations are, for the most part, unbalanced. There will still be 
> some rejection of noise, maybe a good deal. But not as good as if 
> properly terminated.

See my comments above.

>
> I have used twisted pair here with good success, but there is some 
> luck involved if the pairs are not properly terminated.

I have yet to see a situation where twisted pair made things worse, and 
it often solves serious RFI issues. In the pro audio world, it is well 
known, for example, that loudspeaker cables should ALWAYS be twisted 
pair, NEVER parallel wires (zip cord, glorified or not). Sadly, the high 
futility folks have never learned that, and most hams haven't either. 
Audio power amps use feedback around the output stage to reduce 
distortion, a technique first developed more than 100 years ago (by Bell 
Labs, I think). RF on the speaker cable will couple via that feedback 
network to the input of the audio output stage. Replacing the zip cord 
with twisted pair is an effective fix.

Nearly 40 years ago, Prof R. A. Greiner at U of Wis published an AES 
paper that showed that for all practical purposes, the only thing that 
matters with loudspeaker cable is DC resistance, and that lower is 
better.  The sole exception was with a VERY rare type of loudspeaker 
that had a very low impedance at high audio frequencies. The impedance 
of 99.99% of loudspeakers increases rapidly with increasing frequency.  
His paper can be found in the Journal of AES in any decent university's 
technical library. And yes, one of the cables he considered was sold 
under the Kimber name, which was mfd on the concept of litz wire. :)

73, Jim K9YC