Calibration of HF Spectrum Analyzer

[sandy.bridge]

Hey guys,

Cannot seem to find the answer to my question within the Owners Manual for our high-frequency spectrum analyzer. When performing the 2-port calibration, it gives the option of testing in forward or reverse, which is a function of the test cable's orientation (male or female). Is it the male orientation that would be considered forward or reverse, and why exactly?

 

[the_emi_guy]

Are you talking about a network analyzer?

What is the make & model?

 

[f95toli]

Also, by "forward" and "reverse" I assume you mean S12 or S21.

Regardless, the orientation of the connectors does not matter; and there is no standard when it comes to which one is considered "forward".

 

[sophiecentaur]

Calibration of this sort of thing is usually done with standard terminations, mis-matches and attenuators, afaik. You can buy these.

 

[Okefenokee]

It's been years since I used a network analyser but I remember that you perform the calibration on each cable that will connect to the device.

Let's say for example that you are going to test an amplifier. You'll have two coaxial cables from the analyser connecting to the input and output of the amplifier. For the calibration you simply loop the amp's input cable from the output of the analyser back to the input of the analyser (not the amp) then run forward calibration. Next you connect the cable that will be on the amp's output and run backward calibration.

This tells the analyser about the properties of the cables that will be connected to the amp. Specifically, it learns about the phase length and attenuation of the cables so that it can correct for this. Now when you run the test on the amp the analyser will eliminate the effects of the cables and only report the scattering parameters of the amplifier in question.

Some devices are single-ended. An antenna is a good example. In this case your analyser should have a different procedure where you connect standard loads to the end of your cable. It should be a short circuit, open circuit, and standard load. The standard load will be 50-Ohm for a 50-Ohm cable, 75-Ohm for a 75-Ohm cable, and so on. The calibration will again tell the device about the properties of the cable so that it can be eliminated from the results.

edit:

I think analysers use the standard load calibration on both cables used for 2-port devices. The forward calibration is for the cable that goes to the input of the device being tested and the back calibration should be for the output the device.

 

[sophiecentaur]

Another thing you need is an adjustable compensating length of cable in the reference path to take your reference plane to where you want the actual measurement to refer. (To 'unscrew' the phase rotation that unequal lengths will give you.) Without this, you will have no idea whether the mismatches are capacitative or inductive and any phase change may be swamped by cable delay.