Reading time ( words)
In my October column, we looked at the importance of properly terminating cables even at low frequencies and also showed how much detail we can lose in PDN measurements when we use bad-quality cables. In this column we will analyze a step further the shield in our cables.
As a reminder, Figure 1 shows the noise pickup through the shield of two different cables. Data was captured with a 1-GHz bandwidth oscilloscope. Both cables had 50-ohm factory-mounted BNC connectors at their ends. The oscilloscope input was switched to 50-ohm input impedance and the cables were attached to a DUT with a BNC connector. The DUT (a small DC-DC converter evaluation board) was turned off, generating no signal on its own. The noise we see on the oscilloscope is the signal pickup of the cable from the air.
Figure 1: Voltage measured at the DUT BNC connector with two different cables, with the DUT turned off.
The signal of more than 8 mVpp that we see with one of the cables was consistent, not only occasional short bursts. I live in a residential area in the woods, with no visible industrial plant or antenna tower nearby. It might be interesting to track down the source of the noise, but from my perspective it is more useful to look at the differences in the cable braid constructions.
Flexible coaxial cables have conductive foil, braided wire or a combination of these serving as the outer conductor, which also serves as shield in single-ended cables. Braided shields come in single or multiple layers of wire mesh. The cables used for Figure 1 have single-braided shield with no foil.
Read the full column here.
Editor's Note: This column originally appeared in the December 2013 issue of The PCB Design Magazine.