Generally there isn't much common mode from a device that only has two terminals (i.e., the power cable and no other connections), but it can still appear, as the device itself has capacitance to free space, thus there is a return path for CM at high frequencies.
If there are additional connections, then not only can there be unbalanced currents among them (and therefore CM back to the LISN), but external connections should be terminated in a similar way, e.g. telecom pairs through a CDN (see for example CISPR 22).
There aren't many applications for a single terminal device, but it might happen in automotive components for example, where one power wire goes out, and return currents flow through the chassis -- in the lab, the ground plane. This would only need one LISN port, by definition (there's no point in connecting a port to the reference plane, it's defined as zero).
Generally, short or permanent connections between components of the EUT should be laid out in a representative fashion. Maybe they'll act as stubs and affect the EMI profile of the EUT, who knows; hence, "representative". Some standards do perform measurements on such cabling, adding loading ferrites to trim the impedance, using current clamps and inductive or capacitive couplers to read emissions or inject noise...
The assumption underlying all of this, is that the system can be analyzed as a network between a number of ports. The LISN is an interface, joining an endpoint of the EUT's network, to a source, which might be passive (a termination resistor say), or a transmitter or receiver. The response across multiple ports may or may not be relevant (NM on wire(s); CM/DM on wires in a cable; between cables; etc.). The point is, everything required is available on ports, so that whatever the desired analysis is, it can be performed (e.g., using a CM/DM splitter transformer).
Tim
Home
Help
Search
Login
Register
