EMC testing for conducted susceptibility

Hey man, too many acronyms.

Are you injecting into a LISN (Line Impedance Stabilisation Network)?
What frequencies are you looking at? Are you looking at ESD (electrostatic discharge) which is generally far worse than direct conducted interferrence?
I'm not sure if a standard sig gen will give you the amplitudes and spikes you need for decent testing, or do you have some exotice waveform generator?

Anyhow good luck..it's a pig to get meaningfull results without a shed load of equipment.
I ended up using an old Piezo gas ignitor to simulate conducted faults we were finding in commercial laundry sites.

Del

you gave a very nice explanation , this is something which causes me confusion ,In a company's brochure regarding the calibration setup of a BCI probe they attached two 150 ohm of impedances on each side of the calibration jig so that a 300 ohm impedance loop is completed, what is the purpose of this impedance?

while during normal testing they dont use the impedances ?Is that to match the impedance of the CDN method ?

In my earlier response, I alluded to the fact that going into detail could turn this thing into a book chapter. Well, you were warned.

The CDN technique is the desired approach in EN61000-4-6. When a CDN isn't available, then you default to the BCI technique. The CDN, as noted earlier, directly injects the rf potential onto the cable-under-test (hereinafter, CUT). The CDN inserts an rf impedance into the CUT that is very high relative to 150 Ohms, and that impedance is between the CDN and the auxiliary equipment (AE) end of the cable. The reason is to ensure the AE end of the cable doesn't load the rf potential you are trying to inject into the equipment-under-test (hereinafter, EUT) end of the cable.

When no CDN is available, and the BCI clamp injects, there is no control over the cable impedance. If the cable impedance were very low, lots of current would flow relative to the same clamp drive into a high impedance cable. If the cable impedance were very high, then no current would flow regardless of how much power you applied to the clamp. This problem is controlled by pre-calibrating the required forward power into the clamp when the clamp is placed in the 300 Ohm circuit you reference (a short section of transmission line terminated in 150 Ohms on both ends). Now you can see that when you have recorded the number of Watts or dBm that drive the specification-level current into the 300 Ohm loop, that if the CUT presents much more or less than 300 Ohms, the CUT current will be different than the precalibrated current. If the current is less, you simply ensure you are driving at the precalibrated power; this ensures you are injected adequate rf potential. If the current approaches the limit before your clamp drive gets to the pre-calibrated level, you don't increase the clamp drive power beyond that which generates either the specification-level current, or 6 dB above that, depending on the specification.

Thank you i enjoyed this technical answer and a very pragmatic one. But i would like to ask one thing more, how can we be so sure that how will the CUT have an impedance matching to the BCI , it is going to vary so by caibrating to a 300 ohm impedance arent we just assuming too much or is has it been practically measured that all commercial cables have a impedance around 300?

Sorry a bit off topic what is so different in LIa SN than CDN ? and what can LISN be used for specifically?

thank you a lot sir

Your implicit comment is correct: The EUT impedance is not necessarily 300 Ohms. It is, in fact, what it is. Nevertheless, the power limit must be set, and an impedance-matched transmission line is the way to do it (150 Ohms on each end of short line). By mandating a particular way of calibrating maximum clamp power, a standard is established that everyone can understand and design to. Every cable conducted rf susceptibility test, regardless of origin, has the same type of requirement. For automotive, aerospace and military applications, the impedance at the ends of the transmission line is 50, not 150 Ohms, but the principle is precisely the same.

LISN vs. CDN. A very long time ago, the difference wasn't much. But today, a LISN is used exclusively (to my knowledge) to make emissions measurements only, whereas a CDN is used to inject signals. Further, LISNs are designed to be placed in series with a power bus, whether that be mains power, 28 Vdc, or others. CDNs can in principle be designed for any type of cable, be it mains power, 28 Vdc, a shielded multi-conductor bundle, an unshielded multi-conductor bundle, pretty much anything out here. The difference is that the CDN inserts common mode impedance, thus not affecting the intentional signal on the CUT, whereas a LISN inserts differential mode impedance. Therefore a LISN can only be inserted in a low frequency line like power, where the insertion of a few tens to a few hundreds of microhenries has no effect on power transmission.