Chapter 11
Common Mode Choke Design
The selection of the core material and size is the most important part of the design. The size has been arrived at by the transformer design parameters. Normally, the common mode choke shall be the same size or smaller than the transformer. The volume may vary by the changing the height of the core.
The core material must be able to provide a high degree of impedance in the frequency bandwidth of operation in order to suppress the spurious EMI noise. The best material for this would be in the Nickel-Zinc ferrite family and have a permeability of between 450 and 1700.
To be consistent, wind the same size wire as used on the transformer, bifilar, over 320 degrees of the selected core. However, this is not necessarily set in concrete. The wire size may have to be changed to achieve the best parasitic parameters. The number of turns required to fill this traverse shall be recorded as the maximum that can fit on this core.
Now, data has to be acquired for this coil using a waveform analyzer with an impedance test adapter designed for that analyzer. The desired frequency range shall be between 100Khz and 1,8Ghz.
Starting with the maximum number of turns as determined above, the coils are first tested for Impedance over the above frequency bandwidth using the impedance test adapter. The impedance is recorded for each chosen frequency over this bandwidth.
The above procedure is repeated after stripping one turn bifilar off the core. This is repeated until about 9 turns remain on the core. All of this data can then be input into a spreadsheet so that the impedance peaks are easily viewed.
It is a good idea to perform this procedure on various materials and sizes so that the peaks are determined for amplitude and frequency. In this way, it is easier to tailor the coil to the EMI problem areas. An example of this layout is provided on the available CD described separately at the end of this book.
Before assembly, the coils will have to be coated with a buffering agent to prevent magnetostriction of the magnetic material. This will affect the distributed and coupling capacitance by lowering the resonant frequency. Once the magnitude of this effect is determined, changes to the winding turns or winding geometry may be made to compensate for this variation. The attenuation may not be as optimum as first determined, however. Alternate materials may then have to be reviewed to see if, in fact, this is the best one to use.
Once the common mode choke has been wound, it is then connected to the output of the transformer and the common mode test is run on the combination to determine the attenuation at the various frequencies of concern.