USING THE NANOVNA AND SAVER TO MEASURE CM ATTENUATION THROUGH CMCs
|
On Mon, Feb 1, 2021 at 02:00 PM, David Eckhardt wrote:
I promised this, so here it is. See the attachment for a procedure toDave, The measurement of common mode attenuation as it pertains to baluns and RF chokes has been discussed in this group several times and there is a lot of good info in those posts. Reading through these and other excellent references on the Internet one can see the two main areas of concern... - How to measure common mode impedance accurately - Does measuring common mode impedance reflect what happens to common mode current reduction in an antenna system? Let me start with the latter. Several well-known authors on this subject (Tom Roach (W8JI), Jim Brown (K9YC), Owen Duffy(VK1OD), Jeff Anderson (K6JCA) and Steve Hunt (G3TXQ -SK) all stress that knowing the magnitude of the impedance |Z| is not a useful indicator of how well a balun or RF choke will reduce common mode current. As a minimum we need to know the complex impedance R+/-jX and even this only gives a subjective answer. If the choking impedance is primarily reactive the common mode current could get worse depending on the feedpoint impedance of the antenna. What is effective is if the resistance is high and how high is a subject of debate. What is generally accepted is that if common mode impedance |Z| is less than 500 ohms it will be a poor solution. Also there is no direct relationship between the common mode impedance of the RF choke and balun and the reduction in common mode current that would be measured when installed in an antenna system. The second issue is how to measure common mode impedance accurately. It is relatively easy to determine the magnitude of the impedance |Z| using S21 but even that requires careful jig construction especially if we want to make measurements higher than HF. The complex impedance (R+/-jX) can be determined by several methods: S11 or S21. The problem with the S21 measurement method is that it requires additional calculations to determine R+/-jX and the NanoVNA does not have the hardware capability to do this accurately. However S11 measurements can be made on the NanoVNA to determine R+/-jX but the estimate gets poor if we go much beyond 1000 ohms. Just like S21 measurements the test fixture should be built to minimize parasitics and be calibrated with suitable loads. Rather than repeat what the authors above have written on this subject I suggest readers take a look at this post by K6JCA which has a lot of theoretical background, measurements and an extensive reference list. Roger |
|
On Tue, Feb 2, 2021 at 02:34 PM, Dragan Milivojevic wrote:
In order to do the S21 series method of determine complex impedance you need the Port 1 (CH0) and Port 2 (CH1) impedance to be very close to 50 ohms. You also need to be able to do 10/12 point error correction which the nanoVNA cannot do. Both are necessary in order to get an accurate S21 phase measurement. VNA's like Agilent etc can do the measurement because they are bidirectional. Kurt Poulsen described the issue in this post... /g/nanovna-users/message/13075?p=,,,20,0,0,0::Created,,s21+series+method,20,2,0,73384240 I have tried using attenuators on CH0 and CH1 to overcome the 50 ohm issue but the phase was still off even though I carefully constructed test jigs to minimize parasitic capacitance. You can easily see the problem if you make a test jig that measures a 1000 ohm SMD resistor which has minimal reactance at HF frequencies. The S21 magnitude will be accurate but the S21 phase will not be 0 so when you do the calculations you get a reactance that is not there... Roger |
Gary W9TD
Roger Need