A LISN (line impedance stabilization network) is the coupling device that makes conducted emission measurements possible and reproducible. Inserted between the AC mains and the equipment under test, it does three jobs at once: it presents a standardized 50 ohm / 50 uH impedance to the EUT, it blocks RF noise arriving from the building’s mains, and it delivers the EUT’s own noise voltage to the EMI receiver through a 50 ohm measurement port. Without a LISN, the same product would produce different conducted emission readings in every laboratory.

What problem does a LISN solve?

Conducted emission limits - such as FCC Section 15.107 or the CISPR 32 mains-port limits, both covering 150 kHz to 30 MHz - are voltage limits in dBuV. But the noise voltage a product develops on its power cable depends on the impedance it sees looking into the mains, and real mains impedance varies with the building, the wiring length and the time of day. A measurement made against an undefined impedance is not comparable between labs, or even repeatable in one lab. CISPR 16-1-2 and ANSI C63.4 therefore require the measurement to be made against a defined artificial network, making the LISN as fundamental to conducted testing as the antenna is to radiated testing described in EMC testing basics.

How does a LISN work?

Per conductor, a LISN is a low-pass/high-pass splitter built from an inductor and capacitors:

1. Stable impedance. A series 50 uH inductor (with associated resistors and capacitors) shapes the impedance the EUT sees so that, across 150 kHz to 30 MHz, it follows the standardized 50 ohm / 50 uH curve regardless of what the local grid looks like.
2. Isolation from mains noise. The same inductor, together with filtering toward the grid side, blocks RF disturbances already present on the building mains from reaching the measurement, so the receiver sees the EUT’s noise and not the neighborhood’s.
3. Measurement port. A coupling capacitor taps the RF voltage on each mains conductor and presents it to a 50 ohm coaxial port. The EMI receiver measures this voltage with quasi-peak and average detectors; a switch or separate ports select line or neutral, and the higher reading is compared against the limit.

Mains current flows through the LISN continuously, so units are rated for the EUT’s supply current (commonly 16 A, 32 A or more), and the measurement port includes protection because switching transients from the EUT can otherwise damage the receiver front end. Voltage division inside the network is accounted for by a calibrated insertion loss (typically around 10 dB at the low band edge for the standard V-network, falling with frequency) that the receiver software corrects.

When is a LISN required?

Whenever a standard sets conducted emission limits on a power port. That includes FCC Part 15 Subpart B (Section 15.107) for the US, described on our FCC Part 15 page, CISPR 32 for multimedia equipment, CISPR 11 for ISM equipment, and CISPR 14-1 for household appliances. It is equally essential in pre-compliance work: a benchtop LISN plus a spectrum analyzer with a transient limiter is one of the cheapest, most predictive EMC tools a design team can own, since conducted results correlate well between a bench setup and an accredited lab.

What types of LISN exist?

• V-network (V-LISN) - the standard type for AC mains, measuring each conductor’s voltage to ground (an unsymmetric measurement). The 50 ohm / 50 uH (+5 ohm) V-network of CISPR 16-1-2 and ANSI C63.4 is the default for commercial testing.
• Delta-network - measures the symmetric (line-to-line) and asymmetric components separately; used in specialized measurements.
• T-network - used on balanced signal and telecom lines rather than power lines.
• 5 uH automotive network - the CISPR 25 artificial network for DC-supplied vehicle components, with a 5 uH inductance matching short-harness impedance.

The terms used here - conducted emissions, common mode, quasi-peak, dBuV - are defined in the EMC glossary. For where conducted testing sits in a full compliance campaign, see EMC testing basics.