Re: NEC-LIST: Coupling between cables

Hello,

I suppose that you're referring to coaxial cables, used for EMI
testing in a Faraday cage, and that you wish to separate the effects
of cables from that of the DUT. Correct?

In that case the effect of the mode distribution of the testing cavity
are of secondary importance to the cables themselves. The resonances
of the room will however result in frequency selective
measurements. Try calculating them to have a rough idea of where
you'll start having problems. As a general rule, I'd keep cables as
close to the walls as possible.

Coupling will occur when there are common mode (unbalanced) currents
present on the cable. These currents can be caused by :

1) Unbalanced source and load. This is where the configuration of the
DUT's I/O circuits become important. (Balun quality, grounding
path). At one end of the cable would be the DUT, which you're trying
to fix, whereas at the other one are I/O devices, which you don't want
to be part of the test. 2) The cables themselves, through a parameter
called "transfer impedance" related to shielding.

Once you have determined the differential to common mode conversion
factor of the system, you could then use NEC2 to estimate the current
transfer between cables, modelled as simple conductors. Modeling the
room with WGM will only be feasible at low frequencies, mostly below
the bands of interest in EMI testing.

Testing the cables themselves with a network analyzer and a current
probe is also interesting.

Regards,

Alexandre
Received on Mon Nov 06 2000 - 05:31:02 EST