Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Offline SMPS EMC scans: Why are "common mode only" scans so expensive?

Status
Not open for further replies.
T

treez

Guest
Hi,
We find, when we go to get conducted EMC scans (of <200W offline SMPS’s) done at EMC test places in UK, they want big money to do individual common mode or differential mode conducted scans.
So we end up with only having EMC scans which are a mixture of the common mode and diff mode disturbances. (my company can't pay for anything better).

We realise that to get a conducted , "common mode only" scan, we need to have the EMC scan done with a splitter in place, and the EMC receiver has to then do Magnitude spectral measurments.

However, we also believe that an even easier way to get “common mode only” conducted scans, is to use the EMI receiver to measure live and neutral simultaneously…and then post-process tha scan data. I am certain that our EMC test house does this anyway, because their EMI receiver (network analyser) always does just one scan, but they give us separate live and neutral scans from the one scan…therefore, they must be scanning live and neutral simultaneously………their EMI receiver goes straight into a computer, so I don’t see why they want to charge us so much more for “common mode only” scans?
Do you know why “common mode only” scans are so much more expensive?
 

Re: Offline SMPS EMC scans: Why are &amp;quot;common mode only&amp;quot; scans so expensive?

CM scans should not be more expensive - a simple Tx can be used to subtract the DM and give 2x CM on the spec ann - often this is very instructive - as is the DM.

- - - Updated - - -

http://www.hottconsultants.com/techtips/CM_vs_DM Conducted_Emission.html

- - - Updated - - -

http://www.compliance-club.com/archive/keitharmstrong/emc_testing2.html

part 2.1.8 for the Tx structure - obviously wide band on a toroid ...
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Thanks,
Yes, we need to actually take in our own splitter transformer and ask them to connect it up to the LISN outputs upstream of their Network analyser.

Does anyone know of an offtheshelf splitter transformer that’s good for this use?


I looked through minicircuits.com and microcircuits.com and they have nothing.
The other problem is that we don’t know what kind of RF connections they are using. I suppose we will have to ask them, but they have always pretended not to know what we were talking about when we previously asked them about splitting the LISN output into CM and DM.
The problem with us winding our own splitter transformer on a torroid is that we may end up with too much magnetising current and thus distort the signals. Alternatively, we may get too much interwinding capacitance and again distort the signals. We don’t have a network analyser to use to check against such ills of RF design.
The following “LISN MATE” device…

LISN MATE
https://www.tekbox.com/product/TBLM1_LISN_Mate_Manual.pdf

…..splits out CM and DM, but it needs there to be two LISNs, and the test house we go to has a fixed set-up, with just one LISN.

In the following diagram….
https://imgur.com/dc7mAyj

…we assume that the CM output is referenced to earth ground?
..and that the DM output is just the output of the secondary terminals?

I think the problem for us will be insertion loss, but at least we will get an idea of what proportion of the failure frequencys are CM and DM….that will really help.

I must admit it amazes me in this day and age that the computerised Network Analysers at EMC test-houses can’t measure live and neutral at the same time and thus give a separate CM scan plot and a separate DM scan plot…without needing a splitter to be put in there.
 

However, we also believe that an even easier way to get “common mode only” conducted scans, is to use the EMI receiver to measure live and neutral simultaneously…and then post-process tha scan data.
Once the spectrum analyzer converts the signals into magnitude plots you can't combine them to distinguish CM and DM components. It has to be done with an analog combiner before the spectrum analyzer, like the one you linked to.

Strangely, most LISNs only allow you to connect to either line or neutral at a time, so you need to jury rig to LISNs to do this measurement. Maybe your test house has one with two outputs.
 

It's of course possible. You can use a matched resistive power divider/combiner to derive CM, and a balun or a 180°-coupler to extract DM. You should consider that the LISN isn't designed to get calibrated measurements in this operation, phase errors might affect the result.

The LISN is designed to make measurements according to EMI standards with calibrated accuracy, separate CM and DM measurements are not required by usually applicable regulations.

Separate measurements may be interesting for development and EMI analysis purposes, there's no problem to get it as discussed above.
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
It's of course possible. You can use a matched resistive power divider/combiner to derive CM, and a balun or a 180°-coupler to extract DM. You should consider that the LISN isn't designed to get calibrated measurements in this operation, phase errors might affect the result.
Thanks, its dawning on me that if we do this, we are still not going to actually know what proportion of the failure frequency on the official conducted scan is actually common mode , and what proportion is DM. This is because we will of course, have measured each individually with a different technique, each one having its own particular insertion loss.

The splitter transformer in Fig 2.1.8 of the following….
http://www.compliance-club.com/archive/keitharmstrong/emc_testing2.html
…appears a good way to measure CM and DM separately, but with roughly the same insertion loss.

Do you know if these are available to buy off the shelf?
Also, we are thinking that a good way to get a DM only scan (for our offline flyback SMPS) is to actually have a separate PCB comprising five different sized common mode chokes , (to cover from 150khz to 30MHz) and then connect this PCB upstream of the SMPS being EMC scanned. The common mode chokes would be wound to have minimal leakage inductance. This would drastically reduce common mode emissions and thus give a more pure DM only scan. Do you think this is a good idea?
 

Yes - fitting a large amount of CM toroids will leave only DM
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Re: Offline SMPS EMC scans: Why are &quot;common mode only&quot; scans so expensive?

Thanks, also , i believe shorting the primary and secondary grounds of the Flyback SMPS transformer cuts down common mode emissions...
(since in a non isolated Flyback SMPS, the capacitive coupling to the secondary doesnt matter too much because the primary and secondary grounds are shorted together, and any noise coupled to the secondary will just get shorted back to the primary anyway).
Would you agree?

- - - Updated - - -

Also, the web is a little sparse on tips for winding torroidal common mode chokes for low leakage.
I believe that the best way for low leakage is…..
1…Wind both coils bifilar.
2…..make sure that the turns are not bunched up, but spread evenly round the circumference.
Do you know of any more tips?
 
Last edited by a moderator:

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top