Issue link: https://iconnect007.uberflip.com/i/275965
62 The PCB Design Magazine • March 2014 SMA (transmission), and compares the measured received sine waves to the injected sine waves to characterize reflection (e.g., S 11 ) and transmission (e.g., S 21 ). With this instrument, we can measure the full S matrix of a two-port DUT, though to get the full matrix, we have to manually set up four independent measurements. The instrument comes with open, load and short SMA calibration standards, shown on the lower left in Figure 1. For a nominally symmetrical and recipro- cal DUT with two connections, like our cables, there are only two parameters to measure: re- flection (S 11 ) and transmission (S 21 ). If the DUT is really symmetrical, reflection measured from either connection will be the same (S 11 = S 22 ). Similarly, a reciprocal DUT will give us the same transmission measurement result, regardless which direction we measure it (S 21 = S 12 ). Before any measurement is taken, we have to perform a calibration to the end of the con- necting cables. Figure 2 shows the through cali- bration. The two short beige cables are connect- ed with an SMA through piece, also called SMA slug. After calibration, the residual transmission reading by disconnecting the two cables is < - 80dB, pretty good from a pocket-size instru- ment. The two short beige cables have high quality and therefore we get the same noise floor reading regardless where we open the con- nection; whether we just remove the slug but leave the two cables connected to the VNA or if we remove the two cables as well. To check further the shield quality of our co- quiet power CHECKING CABLE PERFORMANCE WITH VNA continues Figure 2: Through calibration layout. Figure 3: layout to measure the two poor quality cables. Top: Both cables with open end. Bottom: Both cables terminated. axial cables, I connected the two cables under test to the VNA, leaving in place the short high- quality cables shown in Figure 2 that were part of the through calibration.