On Saturday 14 July 2018 14:40:24 Chris Albertson wrote: > On Sat, Jul 14, 2018 at 10:26 AM Gene Heskett <ghesk...@shentel.net> wrote: > > On Saturday 14 July 2018 11:57:10 Jon Elson wrote: > > > > Preferably as a twisted pair so the radiation of noise largely > > cancels at 2x the twist rate distance or more. A tight twist = less > > radiated noise. > > No doubt twisting helps, but I have not seen a derivation of the above > 2X the twist rule. > You could say thats a rule derived from my own observations on a scope, noise coupling is usually less than 10% of that measured when the two twisted pairs cross each other at right angles when the separation is about 2x the twist rate, and should not be observable without very careful scope trigger settings by the time the separation is 5x the twist rate.
> On the other hand if you simply use zip cord, with parallel conductors > the equation for the electric field on the plane perpendicular to the > current is easy to derive. (Note that zip cord and a tightly waisted > pair are identical if you look only on the "cut plane") The 2D > electric field is simply the field created by one wire (simple > inverse square law) added on the other and it basically goes to near > zero after 8X the center to center distance of the wires. I always > figured that twisting was just a good way to minimize the center to > center distance. It does better than that since the current going out is matched by the opposing current coming back (if there are no loops), and they rather quickly cancel each other out. > But it could work out that 2X the twist pitch work > s out the same > because small wires have both a higher twist and smaller center to > center distance. > > > I wonder if the twist actually does anything other then just hold the > pair of wires closer together? I think it might but I have never > seen a mathematical derivation. The 2D slice case is easy using only > high school level math but a full analysis of the 3D cable might be > much harder. I think the old ARRL "Ham Book" has a formula for that. I was going to see about it, but looking up at the shelves above me, all I see ATM is an old ITT #5, and I can't recall if its in there someplace in its 1k+ pages or not. > It might be that the twisting is just a mechanical thing and a way to > ensure that wire pairs in a cable bundle are never parallel with each > other and can't capacitively couple. > > Has anyone tried to measure the inductance of a twisted pair. Then > twist it more and re-measure. I don't find any difference at all. > But perhaps My measurement technique is wrong. > > My conclusion is that we twist wires only for two reasons (1) to keep > then closely spaced and (2) to make the path "bumpy" so it is > impossible to lay parallel with anything else. So the effect is > mechanical not electrical. > > I could be wrong, but I've never seen mathematical proof. > > Also when you think about voltages on the grounded return wires, some > people do forget that impedance is a complex number, we are not > dealing with a special case of a constant DC voltage and a resistive > load. Nope, and any current flowing out is matched by a current coming back, and when they add at a distance, they are essentially canceling because one matches the other but has an opposite sign. > You do either want to run all grounds to one and only one common point > or galvanically isolated the systems that you interconnect. In the > machine tool world many systems choose Galvanic isolation using either > opt-iolators or transformers. For a large distributed system > isolation is easier. And both methods insert bandwidth limitations that can muck up the feedback, playing hell with the Nyquist charts and subsequent stability margins. -- Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
