Editor’s note: This question was asked in response to Interference Technology’s recent webinar by Keith Armstrong. To view the webinar, click here.
Question: How can we separate CM and DM noise by oscilloscope?
Answer: Another question that is not associated with the subject of my webinar!
The only two ways I know, whether using oscilloscopes or spectrum analyzers, are:
a) Use current monitoring clamps, such as those made by Fischer Communications Corp.
They aren’t very costly, and by clipping them around a cable (or wire bundle) that contains all of the send and return conductors for wanted signals or power, we then connect the clamp’s output into a spectrum analyzer or oscilloscope to see, and measure, the CM current on that cable (or wire bundle).
Clipping a current monitor clamp around just the send or return conductor for a given signal or power measures all of the DM signal current plus a proportion (half or less) of the CM noise current. Often, the CM noise is such a tiny fraction (0.1% or less) of the DM signal that it cannot even be seen in such a measurement.
Current clamps are available from various manufacturers in various diameters to suit different sizes of cable and bundles, and in various frequency ranges. They are also variously limited in the amount of DM power they can handle when measuring CM noise (e.g. when measuring mains cords).
b) Some Line Impedance Simulation Networks (LISNs) for measuring EMC emissions on mains cords (which other EMC standards might call V-Networks or Artificial Mains Networks, AMNs), are available with additional internal transformers that create the sum and differences of their usual outputs to provide additional CM and DM outputs.
There used to be at least one supplier of a transformer that could be fitted external to a LISN, to convert its usual outputs into DM and CM signals.
Oscilloscopes can sometimes be preferable to spectrum analyzers for analyzing emissions, because they directly show the time relationships in the noise waveforms and so can make it easier to determine which circuit activities are creating them.
We can of course do the same type of analysis with spectrum analyzers, but it requires identifying the repetitious frequency spacing typical of a harmonic structure, then trying to determine its fundamental frequency and hence the circuit operation that is the cause of the emissions.
Both methods have their advantages and disadvantages, especially when there are multiple potential fundamental frequencies (e.g. several different digital clocks that are not spread-spectrum, see Q6 above) all contributing to a general mish-mash of noise.
In the case of a general noise mish-mash, it can help to observe using an oscilloscope, triggering the ‘scope of each suspect clock frequency in turn. With the trigger settings adjusted correctly, if that clock is a source of noise its contribution to the mish-mash will stabilize on the display, leaving all the other noises untriggered and so just a blur. Going around all the clocks, triggering on each in turn, can help establish which circuits and devices are causing the most emissions.
-Keith Armstrong
