Hello,

I downloaded the VNA uncertainty measurement calculator at http://www.agilent.com/metrology/download7.shtml and am now trying to figure out how to correctly apply it. In my specific case, I have a two port circuit that was measured on an E5071B VNA using the 85033E cal kit. I figured out how to enter this information into the calculator spreadsheet. However, I'm not sure about how to apply it to my measurements.

I assume I go to the curves and extract the specific S11m,S11p,S21m,S21p uncertainty. Now, how do I apply these data to my (E5071B) measurements? Do I add these uncertainties to my specific measurements. For example, if my actual S11 magnitude measurement is S11ma (magnitude actual) and the uncertainty is S11mu (magnitude uncertainty), is my measurement uncertainty equal to S11ma +/- S11mu? Same thing for phase uncertainty and the transmission coefficient?

I also want to calculate the dissipative loss of the highly reflective (into 50-ohms) 2-port by using the formula Loss=|S21|^2 / ( 1 - |S11|^2) . Will the uncertainty analysis provide a useful estimate for loss? Is there any correlation between the S11 and S21 uncertainty?

Thanks

I downloaded the VNA uncertainty measurement calculator at http://www.agilent.com/metrology/download7.shtml and am now trying to figure out how to correctly apply it. In my specific case, I have a two port circuit that was measured on an E5071B VNA using the 85033E cal kit. I figured out how to enter this information into the calculator spreadsheet. However, I'm not sure about how to apply it to my measurements.

I assume I go to the curves and extract the specific S11m,S11p,S21m,S21p uncertainty. Now, how do I apply these data to my (E5071B) measurements? Do I add these uncertainties to my specific measurements. For example, if my actual S11 magnitude measurement is S11ma (magnitude actual) and the uncertainty is S11mu (magnitude uncertainty), is my measurement uncertainty equal to S11ma +/- S11mu? Same thing for phase uncertainty and the transmission coefficient?

I also want to calculate the dissipative loss of the highly reflective (into 50-ohms) 2-port by using the formula Loss=|S21|^2 / ( 1 - |S11|^2) . Will the uncertainty analysis provide a useful estimate for loss? Is there any correlation between the S11 and S21 uncertainty?

Thanks

So, for S21 uncertainty you add the dB uncertainty +- to the S21 measuerment, same as phase. BUT, the uncertainty is really something like a coverage factor of 2.77 so the standard deviation is quite a bit smaller. If you are adding the uncertainty of more than 1 thing, it is normal to turn the spec into a standard deviation, then rss the standard deviation, then turn it back in to a coverage (2.77 std dev is 99% coverage, I believe).

For S11 uncertainty, you take the S11 vector, and add the S11 uncertainty to it. Say the S11 is .98; and the uncertainty (which changes as a function of reflection) is 0.03 linear at your reflecdtion coefficient. Then the S11 max is 1.01 and the S11 min is 0.95: This means (in dB) that the S11 range would be: -.445 dB> -.175.dB> +0.086dB (min: nom: max) ; but in your formula, S11>1 is a no-no so such a case would mean you can't say anything about loss.

Phase for S11 forms as the arc tan (unc/S11) or (in this case atan( 0.03/.98)=1.75 degrees. Notice for a small reflection, one that is about the same size as the uncertainty, the phase uncertainty goes to +-180 degrees!