• 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.

Mains conducted EMC test with picoscope?

Status
Not open for further replies.

JimB

Super Moderator
Most Helpful Member
No, not sensitive enough for meaningful measurements.

Look at the sweep of your emissions as done by the test house.
The midpoint of the vertical scale is 60dBuV, in other words 1mV.
The most sensitive range of the PicoScope appears the be 50mV.

You need a proper spectrum analyser, not some USB scope with an FFT function.
I suggest that you look at the various second hand test equipment suppliers websites.

JimB
 

Flyback

Well-Known Member
Thanks, i see your point, but even if its only extremely rough we need something, since we have failed mains EMC conducted emissions as per the attached EMC conducted scan.
We cannot understand why we have failed the EMC conducted test...because our led streetlight only switches 10 times per mains half cycle, and these are gentle switching transitions.

The mains input current to our product is as in the attached...as you can see, it shouldnt have failed this badly.

We do not want to pay for a spectrum analyser because our product doesnt contain a SMPS, so we shouldnt really be failing EMC, so we shouldnt really need to do EMC scans.........we are baffled why we have failed EMC when we only have some 10 gentle switching transitions per mains half cycle, whereas an SMPS has millions of much more noisy switching transition per mains half cycle.
 

Attachments

kubeek

Well-Known Member
Most Helpful Member
What exactly do you call gentle transitions? What I see are wild oscillations.
Anyway. without at least a basic schematic showing the switching element action and input filtering i doubt anyone can give you any sound advice.
 

JimB

Super Moderator
Most Helpful Member
The mains input current to our product is as in the attached...as you can see, it shouldnt have failed this badly.
You are joking!
That is EXACTLY why it is failing so badly.
Narrow spikes and pulses make wideband noise.

JimB
 

Flyback

Well-Known Member
The time base on that input current scan is approx 1ms/div. The “craggy” bits that you see have di/dt’s far less than a normal SMPS's switching transitions. So why is it failing as if it were an SMPS?

Also, as discussed, a normal SMPS switches very noisily and with high di/dt about 100000 times per second……our power supply switches just 1000 times every second, and each of those switching transitions is far less in di/dt than an individual switching transition of an SMPS. So why is it failing, when it has 100 times less switching transitions than a typical SMPS?

Since our switching transitions occur every 1ms, then do you think we stand a good chance of failing the mains emissions at that mains harmonic?, ie the 20th mains harmonic frequency will be a harmonic failure for us?
 
Last edited:

tomizett

Active Member
Surely the results of the test should point you in the right direction (by looking at where in the spectrum you where over the limit)? I'm afraid the plot that you posted means nothing to me, so I cant make any suggestions - although maybe others can. It's really impossible to tell from the current waveform what might be in there, given that the timebase is so low - you'd have to zoom in on one of those spikes and take a closer look.
If you can narrow down the frequency and whether you problem emission is common or differential mode then you have some idea what to look for with a scope connected to the LISN, even without a spectrum analyser. You should then be able to play around and see what circuit modifications reduce or increase those signals.

Regarding the Picoscope, can't you just take the best scope you have in your workshop (I'm assuming you're doing hands-on R&D on this) and use that? I'd have thought you'd want to use the best instrument you have available for such a critical measurement.

Sounds tough though. I've just bought my first book on EMC, and I get the impression that failing those tests can get expensive very quickly. It must be a big (albeit necessary) barrier to bringing new products to market.
 

large_ghostman

Well-Known Member
Most Helpful Member
Fly back you in the UK? What is your budget for a spectrum analyzer?
 

Flyback

Well-Known Member
Yes i am in the UK. We are reluctant to pay much for a spec analyser as our products dont use smps, so shouldnt really need specturm analyser for the emc test.....
 

large_ghostman

Well-Known Member
Most Helpful Member
The test is a requirement, dosnt matter what powers it. pretty much anything powered need to pass the test.

Your pics show why this is one of the few good rules we have, I asked because I have a HP Spectrum Analyzer I might let go of at a good price, but you keep paying for tests and failing if you want to go that route. Think how many years it is going to take by trial and error, how much you will spend on the tests and how many sales you will use. One good thing about it is, it should now start to debunk some of your prejudice against SMPS. Your seeing first hand that theory and fact are not twins.
 

The Electrician

Active Member
You could always rent a spectrum analyzer for a few weeks instead of buying one.

It would be nice if you would show only a single cycle of the waveform instead of five.

The EMI spectrum depends not so much on the number of transitions, but rather on their transition times. What are the rise and fall times of those "gentle" transitions? Show a scope capture of some of them.

I think you may also have another problem. Don't the regulatory requirements place limits on the low order harmonics, such as the 3rd through the 40th, or so?

It looks like your waveform is a stepped approximation to a sine wave with 5 voltage levels per quarter cycle (not counting zero), with the ideal looking something like this:

Sinw1.png

Such a waveform has about 8 to 9 percent THD, probably exceeding the limits on the percentage of certain specific harmonics, such as 3rd, 5th, etc.

You may need to have more steps in the waveform.
 

Flyback

Well-Known Member
Thanks, i attach one mains cycle here..

Supposing you have a 150W offline product which puts a burst of say 30khz decaying amplitude ringing frequency current into the grid, of say 200mA pk-to-pk max, and does this for 200us every 1ms.
What dmage would this do to the grid?

I understand the point about interfering with analog radio's in the area , that use AM or FM modulation, (because the frequency could potentially radiate from miles of national grid electricity supply cable) i also understand about heating due to core and hysteresis losses in the supply system transformers....but what other damage is there to think of?

Imagine that there could be thousands of such products connected to the grid.

Example of mains input ringing current supplied attached

Also, we have one current step every 1ms...so thats 1khz.....which equates to the 50th harmonic, and they only test up to the 40th harmonic of the mains. Do you agree?
 

Attachments

Flyback

Well-Known Member
Such a waveform has about 8 to 9 percent THD, probably exceeding the limits on the percentage of certain specific harmonics, such as 3rd, 5th, etc.
Thats what i first thought, but then the power factor is 0.99+....so it couldnt possibly be failing on those harmonics, or any others.

Another point is that because we dont have an SMPS, we can use iron cored input inductors and get 100mH with a small size.....then our mains input current waveform looks like the attached....and i am sure you would agree this isnt going to fail any mains harmonics whatsoever.

Incidentally , do you know where we can buy iron cored inductors from?, i looked on mag-inc.com and micrometals website , etc etc but couldnt see any.
 

Attachments

Last edited:

large_ghostman

Well-Known Member
Most Helpful Member
This is all semantics really. You shouldnt focus on other products or bother speculating how many bad products are around. Your sole focus is getting YOUR product through the tests.

It is a mute point whether you believe your device should have passed or not, the fact is it failed the test. That should be your only focus, worry about how to fix your problem, rather than waste time comparing it with others.

Sorry to be so blunt flyback, but you tend to get in the pulpit on **** like this. Your a smart guy but your a bit highly strung, pay the shipping both ways and insurance and I will lend you mine. Its a old but very very decent HP spectrum analyzer, I will get the model number later as I cant remember. Its very heavy so be warned. But if it helps you out I would lend it FOC.
 

The Electrician

Active Member
Thats what i first thought, but then the power factor is 0.99+....so it couldnt possibly be failing on those harmonics, or any others.

Another point is that because we dont have an SMPS, we can use iron cored input inductors and get 100mH with a small size.....then our mains input current waveform looks like the attached....and i am sure you would agree this isnt going to fail any mains harmonics whatsoever.

Incidentally , do you know where we can buy iron cored inductors from?, i looked on mag-inc.com and micrometals website , etc etc but couldnt see any.
The waveforms in this image appear to be simulations rather than scope captures; is this the case? Have you got any scope captures?

I think there is more distortion in the green waveform than you think.

Here's a plot of a pure sine wave (blue), and that same sine plus 10% 3rd harmonic (red). Notice how very similar to your green waveform this is:

SinPlus3rd.png
 
Last edited:

kubeek

Well-Known Member
Most Helpful Member
I suggest you get or build a LISN and buy or borrow a spectrum analyzer. Also don´t forget to measure your background before you measure your device so that you are not chasing ghosts.
 

Flyback

Well-Known Member
Thanks The Electrician,

What you say is of great interest. You are absolutely right, I ran the Fourier transform on the simulator and it comes out as just over 10% at the third harmonic. I then ran a confirming LTspice simulation, and it shows the waveform being just like ours when the third harmonic is set to 10% of the fundamental.

Table 2 , Page 7 of the following…….
http://www.epsma.org/PFCver100406_b.pdf

….shows that for Lighting equipment, the third harmonic is allowed to be no more than 2% of the fundamental input current.
This means that our 150W lighting product is a total failure on Harmonics (EN61000-3-2).
I cant understand how its set to 2% though. I mean, this is a tiny figure. The above document shows the harmonic limits being far slacker for non lighting equipment. Why is it so strict for lighting equipment?

We do have scope shots but yes the ones I have here are from the simulation.

Does anyone know why the Harmonic Limits are so ridiculously tight for Lighting Equipment?
 

Flyback

Well-Known Member
Woops, sorry, i am guilty of reading the wrong entry in the Table 2......for the Third Harmonic, in lighting equipment, the limit for the third harmonic is 30*PowerFactor percent of the fundamental...i need to recalculate....the 2% figure is for the 2nd harmonic
 

large_ghostman

Well-Known Member
Most Helpful Member
Thanks The Electrician,

What you say is of great interest. You are absolutely right, I ran the Fourier transform on the simulator and it comes out as just over 10% at the third harmonic. I then ran a confirming LTspice simulation, and it shows the waveform being just like ours when the third harmonic is set to 10% of the fundamental.

Table 2 , Page 7 of the following…….
http://www.epsma.org/PFCver100406_b.pdf

….shows that for Lighting equipment, the third harmonic is allowed to be no more than 2% of the fundamental input current.
This means that our 150W lighting product is a total failure on Harmonics (EN61000-3-2).
I cant understand how its set to 2% though. I mean, this is a tiny figure. The above document shows the harmonic limits being far slacker for non lighting equipment. Why is it so strict for lighting equipment?

We do have scope shots but yes the ones I have here are from the simulation.

Does anyone know why the Harmonic Limits are so ridiculously tight for Lighting Equipment?
Because everybody has lights? I mean if you think about it lights have got to be the most abundant device on the planet. Again though the why dosnt matter does it? Now if no light company was passing the test you would have a point, but plenty do pass the test. So you have to reconsider everything you thought was right about lights.

You have insisted for years your way was the better way, then the guy comes does the test and you fail. You cant say the test is too tight when others pass it, maybe that answers your main question, why do people do X when Y would be better. Now you know why they use X because it gets through the test and Y dosnt.
 

Flyback

Well-Known Member
Yes, sorry , ive just calculated that our third harmonic current is 71mA, and the regulations say that in our case it could be up to 280mA.
I must admit i am a bit worried about this 2nd mains harmonic, as thats only allowed to be 2% of fundamental, so i'll have to go and check that now.
I do apologise for me confusing myself......i am going cross-eyed looking at all these tables and regulatory standards.

By the way, supposing that the 2nd harmonic was 4% instead of 2%, what possible damage could that do to the electricity supply system?
 
Last edited:
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top