Automotive LV LISN

The LISN we're building is specified as the 5µH / 50Ω LISN in CISPR 25. It consist of just an inductor and some capacitors.

The values of R1, R2 and C1 are not too critical, but especially L1 should be sufficently "good". For the capacitor C2, any not too physically big foil or ceramic cap should do. Remember, the bigger (physically!) the capacitor, the lower its resonance frequency gets. Since we're after 150kHz-30MHz, there's no need to get paranoid about this...

The 5µH inductor.

In the previous section, I mentioned that we are after a "good" inductor. For our purposes, we need to consider 2 crucial properties:

• Self Resoncance Frequency (in datasheets often abbreviated by SRF)
• Saturation

The first one is very tricky. A classic LISN is build from air-core inductors. These are quite significant in sice and this are prone to show all of the parasitic effects you don't want to see.

Using small SMD power inductors, we can get around this. Luckily I had 10 small 2.2µH inductors from unknown brand that weekend I started building this LISN. The Aliexpress listing suggests a rating of 3A, so when in parallel, I get 1.1µH and limiting to 5A will give me 60W at the DC-plug. That's right what you have at the most 12V plugs in modern cars.

Construction

Usually, my go-to enclosure is the Camdenboss CMCB00002. That was not available at the time I designed this. The next best thing I had around were some steel enclosure from RS-Components from their RS-Pro brand.

After laying out everything with some pen I accidentily though of being a whiteboard marker, the holes were cut and everything was ready to get clean again. Well, these enclosures are sturdy and have some weight. The color is nice but any permanent marker will leave significant remains... so be warned... I had to touch everything off with some rattle-can paint.

As shown in the photo above, I cut some length of copper wire, and soldered 5 pairs of 2 2.2µH inducotors from the input to the output. While assembling everyting, I noticed, that I should have measured the impedance of the inductor. That's why you can already see some more things in the photo...

As you can see in the screen shot above, the inductor is far from perfect. It has a parallel resonance at around 25MHz. That's not a problem, since here, the 50Ω from the measurement instrument will define the impedance. I tend to call inductors usable as long as they stay above 500Ω (that's 10x the nominal impedance) so that they do not contribute to the total impedance too much. Looking at the graph suggests the inductor should not limit the performance too much.

Connecting the capacitors and resistors is rather trivial... you'll see them in the photos below in the next section. Just one further comment, I needed to remove 1 pair of inductors to hit the specs in CISPR 25.

Calibration

Calibration is a rather easy task. The only specification is about the impedance at the EUT port. If you look at the impedance numbers, CISPR 25 seems to consider 47Ω as the norm.

This got me into troubles. My termination is done using the regular 50Ω. In addition I have slightly higher nominal inductance (6µH instead of 5µH).

Long story short, I removed a pair of inductors to meet the impedance requirements.

When building the LISN, I already had some EUTs in mind. Both were 12V USB chargers. Hence, I took my cheap and poor single charger from my car, removed all guts and connected a SMA connector between GND and +12V.

This wacky contraption is right next to the LISN in the photo above. Since I wasn't too sure about the performance of this calibration "jig", I opted to add another SMA on the inside right at the connector. My comparison showed some difference at higher frequencies (as expected), but nothing which would make the LISN fail the impedance requiremens if I came up with something better.

As you can see, we have a usable LISN up to 30Mhz. The connection of the measurement port is rather problematic at higher frequencies. If you really need to go up to 100MHz, you should definitly add an attenuator directly at the DC-connector. For my purposes (up to 30MHz), the performance is fine, though.