Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Makeing our own LISN for mains conducted EMC testing

Status
Not open for further replies.

Flyback

Well-Known Member
We wish to make our own cheap Conducted Emissions test kit for our 50-250W UK mains powered LED luminaires. We don’t need perfection. We are trying to make a LISN. The inductor in the LISN is to be 50uH. This inductor is supposed to be air core, made by hand winding over plastic drum . The inductor should feature damping resistance along the coils to reduce ringing due to the inductors self resonant frequency. However, if we just pick say an inductor with a SRF of about 7MHz, then does that mean that our readings of conducted emissions will be OK up to around 5MHz? Also, suppose we want an increased SRF, then can we just solder in series a load of say 3uH inductors till we get to the 50uH value (then the SRF will be higher)

Making a LISN…
**broken link removed**

At the moment we have picked this inductor...
http://uk.farnell.com/coilcraft/pcv-...-10/dp/2457647


The LISN diagram of the above link document (page 2) shows an input filter as well as the LISN. Does the input filter have to have these exact component values. (L2 = 250uF and C2 = 4uF)
Also, it says that the L2 inductors are coupled inductors (common mode choke)….but I thought the LISN was also supposed to be able to measure common mode emissions?
Also, for measuring common mode emissions, I presume the kit has to be able to allow measurement of 100MHz plus? Common mode conducted emissions, as I understand it, are rarely prevalent below about 10MHz?
Also, how do you make a LISN that can allow common mode emissions measurement? Do you have to use a coupling network out of coupled coils?
Also, presumably there are problems when measuring common mode emissions because the signal level will be extremely low….so I presume some kind of broadband RF amplifier will be needed for its measurement?

Also, if we dont make it in a metal box, (instead use plastic) then to what frequency will the LISN be effective up to?
 
To what standard will you be testing you luminaires?

These fixtures are not as easy to make as it sometimes appears. You should strive to match the requirements of the standard as close as possible. If you don't strive for perfection you have a good chance of coming up with an unusable piece of test gear. Based on that, your trick of using an off-the-shelf inductor with a ferrite core probably will not give the desired results. Putting multiple small inductors to series to increase the SRF will only give you a network with multiple SRF's,

You asked the following:

1). However, if we just pick say an inductor with a SRF of about 7MHz, then does that mean that our readings of conducted emissions will be OK up to around 5MHz? Maybe, that depends a lot upon the harmonic content of the EMI and how well you build the LISN. You should use the LISN specified in the standard.

2)Also, for measuring common mode emissions, I presume the kit has to be able to allow measurement of 100MHz plus? Common mode conducted emissions, as I understand it, are rarely prevalent below about 10MHz? If you don't build the LISN per the specification you will not make meaningful measurements.

3). Also, how do you make a LISN that can allow common mode emissions measurement? Do you have to use a coupling network out of coupled coils? I don't knonw specifically. You will be testing according to some standard. You should follow setup, method, and calculations strictly according to the standard or yo won't be able to compare your results with the standard.

4) Also, presumably there are problems when measuring common mode emissions because the signal level will be extremely low….so I presume some kind of broadband RF amplifier will be needed for its measurement? In my experience this is usually done with a very nice (expensive) spectrum analyzer, though in some case a tunable voltmeter can be used. Again, what does the standard say?

5) Also, if we don't make it in a metal box, (instead use plastic) then to what frequency will the LISN be effective up to? No way of telling. That's one reason LISN's are made in metal boxes. If you don't follow the standard you can only get rough results.
 
Last edited:
Thankyou,

We are testing to EN55015. We have failed badly across the conducted band of 150khz to 30MHz. Seems a bit strange given that our 150w offline led driver is not an SMPS based one, but just gently switches in/out led banks as the mains half cycle rises and falls.
We failed with no EMI filter in circuit, because we wanted to see how good/bad it was because we think we should be able to get away with a small filter because we are not using a switch mode....thing is, we cant afford to keep paying the EMC test house every time we modify the filter and need a re-scan.....we need to do something ourselves first as kind of a pre-test.......we have the test house plots for our luminiare with no filter, so we are reckoning that if we build a LISN and get the same plot as the test house, when no filter is there, then we have a good LISN (if it gave the same plot as the test house then it would obviously be good) ....or good enough. We would like to buy a LISN, but cant affird the £1000 for a HM6050.
 
Flyback,
I have a home made lisn and old test equipment that was last calibrated many years ago. (calibration report on clay tablets)
I do not look at "numbers" or db. I take the lab's report, like the one you have. If they say it is 3db too high at 150khz then I find a way to bring down the reading at 150khz with out pushing some other point up much. (I use relative numbers not actual numbers) My test machine might be -5db at 150khz from calibration but I don't care. I just need to move the line down by 3db.
RonS
 
Thanks, the following shows how badly we have failed the EMC scan. -Across the entire band, which seems remarkable when there are a couple of led banks that get gently switched in and out twice every mains half cycle....and no high frequency switching power supply....of course, this scan is with no filtering whatsoever.
 

Attachments

  • EMC scan conducted emissions.jpg
    EMC scan conducted emissions.jpg
    84.3 KB · Views: 209
which seems remarkable when there are a couple of led banks that get gently switched in and out twice every mains half cycle....and no high frequency switching power supply.
If I remember right you are slamming the LEDs into the power line with out any inductance. There might be some real current on the switching edges.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top