I work in a Metrology Lab and perform calibrations on a wide range of power meters. I am a little concerned with some ambiguities I am seeing when I am measuring the "Power Reference" output levels of power meters utilizing the procedure called out in most of the service manuals. Per the procedure I am using a model 432B Power Meter and a model 478A Thermistor Mount, both freshly factory calibrated and certified and a calibrated 34401A DMM. I am comparing this measurement against one I make using a freshly factory calibrated and certified E4419B and several factory calibrated and certified 8480 series power sensors. I just performed these measurements on an E4418B power meter and using the method called out in the manual calculated it to be 0.981 mW which means I would need to adjust the power reference to meet spec. However, when I measure the power ref. with the E4419B and power sensor I get a reading of 1.001 mW whic is almost perfect. This is quite typical of my results on all the power meters I calibrate. Why the desparity between methods and is the alternate method I use acceptable?
There could be a couple of reasons for this difference. But I need to ask one question first. When you measured the E4418B's 1 mW reference power level using the E4419B and its sensor and got 1.001 mw what 1 mW reference source did you use to calibrate the E4419B/sensor combination? Was it the E4419B's reference, or another reference?
I used the E4419Bs Power Reference. I had just received it back from a vendor calibration in which I requested that they optimize the reference as it was my lab standard and I wanted it to be as accurate as possible. This was also the case with the power sensors. I route at least 2 each of the various models I have in the lab as standards back to the vendor each year to have them characterized.
I am not sure what is causing the difference between the measurements. In the way you are using the 432B it should be more accurate than the E4419B that you also used. I am not sure what sensor you are using with the E4419B, however the measurement uncertainty for the E4419B with a E9300A sensor is only 2.072%. The 432A uncertainty is even less than this. So both methods should get you an answer that is very close to each other.
Have you got another method to verify the accuracy of both of these power meters?
Check to ensure the 478A you are using is an actual -H73, -H75, or -H76 option. Only the ones with these options have the required SWR to perform the measurement of a power meter reference. Also, measure the actual resistance of your 432A and cable so you can use that resistance value in the power calculation instead of the nominal 200 ohms.