What is the theory behind using a sliding load to perform the cal on a VNA. I know that it changes the phase and somehow yields for greater return loss measurementsâ€¦but why and how?? Out of the six set points (such as the sliding load in the 85052B kits), what would be the results if I only used four of the set points?? What would the results be if I also performed a set in between each detent making eleven or twelve set points?

Is there a way that I can test my system after a calibration to make sure that it is capable of measuring devices with a return loss of 40dB or greater? If I connect the sliding load back up to the VNA after a calibration then it shows a return loss of 30 to 35 dB although the sliding load has a return loss spec. of 44dB. --When I test the directivity of bridges with very low return loss, we move the sliding load back and forth to find the max. and min. return loss values of a particular frequency of interest. We then use a â€œSignal Separation Chartâ€ to find the â€œEffective Return Lossâ€. Is this method the same if I wanted to measure the return loss of the sliding load to verify the cal of the 8510C or any VNA at higher frequencies? Is there a formula to calculate the â€œEffective Return Lossâ€ of the sliding load? ......Should I be able to re-measure the sliding load if I just used it to cal up my system?

Is there a way that I can test my system after a calibration to make sure that it is capable of measuring devices with a return loss of 40dB or greater? If I connect the sliding load back up to the VNA after a calibration then it shows a return loss of 30 to 35 dB although the sliding load has a return loss spec. of 44dB. --When I test the directivity of bridges with very low return loss, we move the sliding load back and forth to find the max. and min. return loss values of a particular frequency of interest. We then use a â€œSignal Separation Chartâ€ to find the â€œEffective Return Lossâ€. Is this method the same if I wanted to measure the return loss of the sliding load to verify the cal of the 8510C or any VNA at higher frequencies? Is there a formula to calculate the â€œEffective Return Lossâ€ of the sliding load? ......Should I be able to re-measure the sliding load if I just used it to cal up my system?

The return loss spec of a sliding load comes from the quality of the impedance of the airline between where it connects to the test port, and where the load element is. The actual value of the load element is not important, as long as it remains constant as the slide is set.

A sliding load is exactly the same as using several offset loads, where each one is a bit further from the test port. If the airline that separates the load from the test port is nearly perfect, the load will form a circle around the center of the smith chart. Since 3 points define a circle, that would be the minimum number of slides for a sliding load to have, but I think we restrict it to at least 5, which allows a better fitting of the circle.

After you do a calibration, put the sliding load back on and you will see the load is not at the center (use a marker so you don't get confused by the frequency response of the load). As you move the slide, the magnitude should stay exactly the same, but the phase will change. It doesn't really matter if the load is -20 dB or -30 dB, because we use the center of the locus of points describing a circle as the reference impedance. Thus, to verify a sliding load, you must know the value of the airline, which you can measure mechanically, and know the impedance of the coaxial line by direct calculation.

So a more proper term should be "sliding mismatch", "but the wisdom of our ancestors is in the simile; and my unhallowed hands shall not disturb it, or the Country's done for." (Now who know's where that quote's from :?: )

Now to answer all your questions: 3's the minimum, we require 5 and (maybe) ignore more than that (don't remember right now). You can test the system directivity only with an airline of higher quality than what you are testing. If the sliding load is reconnected, and slid, each point should have identical spacing from the center. Any error is residual directivity. Sliding the load back and forth on a bridge moves the mismatch in phase to match the worst and the best to the bridge, so the "perfect" match is somewhere in the middle. You can re-measure a sliding load as described above. Note, if it is the same distance from the center, that means that you will so NO variation in return loss from your sliding the load. Any variation must be residual directivity.