Hi. this will be just a relatively uh, quick follow-up video to my previous mailbag one where um LD from uh Techbox.sk entry pre-compliance testing with one of these uh line impedance stabilization networks or listens, um and a cheap bottom of the range Riol DSA 815 Spectrum analyzer Now Uh, pre-compliance testing is a big deal for um, anyone really any manufacturer manufacturing any electronic product that you want to sell in the commercial Market In pretty much, um, any country in the world at least if you you know care about uh, doing it Le legally and correct and you've probably seen these symbols around this C Mark and this FCC Mark here and in Australia this uh C tick symbol here Now uh, different countries have various standards. The C standard. This is Um EMC Compliance Electromagnetic Uh Conformity Basically in the European uh, Union have that standard in Uh America it's the FCC in Australia It's this CTIC compliance standard.

So depending on where you want sell your product into What markets, um, you may have to test your product against these various standards to get this Mark on your product. and uh, it can cost thousands to tens of thousands of dollars to get your uh product tested. So you've design your widget like this uh Main's plug pack here and you you know you've done all the proper design um, rules and you know what you're doing and you're pretty sure it's not going to radiate much and uh, you know it's not going to be susceptible to outside interference even if you care. Um, and you know you go to get the thing EMC Compliance tested and fail.

You forgot something and it's failed the um very stringent uh requirements or very specific requirements for any one of these particular things. There are more standards than this of course and it depends on the type of product which we won't go into. The details are are enormous, but uh, when you design your product, if you fail that pre compliance testing uh sorry, not pre-compliance proper compliance testing then you'll have to respin your product or figure out what's wrong first and then respin your product and get it tested again. And to do that over and over again on a trial and error basis is a real pain in the ass.

So any smart company will what's do called uh in uh, pre compliance testing and you can do that inhouse with as I said one of these uh line uh, impedance stabilization networks for conducted measurements and a cheap ass Spectrum analyzer which you can get these days like this Ryo one but there are several types of emissions that uh, your product can actually uh emit so we have to talk about those quickly cuz we're only going to look at one today now. I Got this diagram from Uh Wikipedia thank you very much Steel Pillow because it quite nicely illustrates the four different Uh coupling, noise, coupling uh, or emissions coupling mechanisms from your device here which will uh call The Source One Of the major ones, of course, which everyone thinks of, uh, when they think of, you know EMC uh compliance And things like that is what's called radiative emissions. And you're probably familiar with this. You know your electronic product has lots of switching frequencies in there very uh, fast sharp digital signals uh, which emit very, You know Broadband uh radiative.

You know noise, emissions, and things like that. So you know pretty much any product. unless it's completely sealed in a metal can with nothing coming in, nothing coming out, and the metal cans 100% effective. Uh, over the whole frequency range or the desired frequency range, then pretty much your product is going to radiate electromagnetic interference out all those little PCB traces, everything inside acting as little antennas and woof out it comes.

And we're not going to look at this one today, but that is one of that is the major one that everyone thinks about and and is a big deal with EMC compliance. But the other one fewer people know about is uh, conductive emissions and many products will have to be tested for conductive emissions as well. And this is what we'll cover today. There are two others, Of course there's capacitive, uh coupling from your source to your victim over here, as it's called.

And there's also inductive coupling as well. But because inductive and capacitive coupling are nearfield effects, uh, pretty much they are only Uh cover. They're only required, uh to meet some sort of standard in very specific cases for products. So most products will not be tested for these nearfield effects.

but they will be tested for radiate, radiative, uh, interference, and uh, or emissions and conductive emissions. So even if your product was battery powered like this inside a nice diecast metal shazzy like this, and you know it's got the lid sealed on and everything. Well, in theory, um, then you wouldn't have any radiator, uh, emissions, or conducted emissions cuz there's no cables coming in and out of it. But that's not a very useful product is it? Usually there's got to be cutouts for screens and things like that so it can actually radiate signals out.

or in this case, even if we had something like this and it was completely shielded like that. Bingo We've just broken that shield with our BNC here and noise from inside any switching noise can couple onto the signal and onto the coax and bingo out it comes. And that's called conductive emissions cuz it's conducting along the copper cables into and out of your product. And take for example, the Uh plug Main's plug pack here.

If this could, this could be completely metal shielded. Uh, till the cows come home and you know it may not have any radiative, uh, emissions, uh to speak of really. But you've got the Main's cable coming in and you've got the DC power cable coming out. So in that case, if this was completely shielded for radiative stuff, your conductive stuff could kill you and you might fail your compliance testing.

So to do conductive pre-compliance testing, we need one of these uh listens or line impedance stabilization Network and I'll explain why it's called that in a second. This one is from Uh Tech Box and uh, send it in, it is open- Source Hardware Brilliant! And here's the schematic: I'll provide the link down below and you can buy it from them or you can make your own. So what this is designed to do is insert in series with your your power cable coming in and your power. uh comes in here and then this end here.

Plug. Well, this end these two terminals plug into your product and what it does is it. AC couples off the signal into your spectrum analyzer there and this inductive filter here basically provides a fixed impedance. hence the name line Impedance stabilization Network Because basically your supply over here is an unknown impedance.

You don't know what that's going to be. It could vary or whatever. It's no good, especially when you're doing standards testing if your Source varies. So uh, all these compliance testing standards specify a specific 50 ohm Network like this, so this is what it does.

This provides a fixed 50 ohm uh line impedance that you can do a controll testing over um over a specific frequency range which we'll uh see in a minute and it just Taps off the AC signal. And a good listen device like this one will also contain that surge suppression as well. We got some 5V TVs diodes in here to protect our Spectrum analyzer from any surges on the power supply here. and there's a gas discharged tube and a Mau over here, so it's pretty well protected.

Otherwise, you got to be careful when you power up your product. if it's connected to your delicate uh RF input of your spectrum analyzer here, you know how it's always got that uh, you know, maximum uh value. This one's pretty good at uh, you know, 50 Vols at DC maximum, but some of them, you got to be real careful you don't blow your expensive Spectrum analyzer. And there are various Uh standards for these line impedance stabilization networds.

One of them is the Uh Cisper standard here and you can see that uh, the standard actually provides um, what the Uh impedance of this uh Listen network is supposed to be. Here's the upper and lower limits and you can see this one is actually designed to fit. You know fairly much in the middle of that. So this one actually meets the requirements.

So having a standards compliant listen box like this absolutely essential for any Uh conductive emissions pre- compliance testing. Now a proper EMC compliance Uh test facility will have all a proper shielded room to do all this in. They'll thoroughly meet the requirements, they know what they're doing, but if you're doing Simple pre-compliance in your lab, then uh, you know this is a basic uh, best practice way to do it and the standards do actually uh specify this and dimensions and things like that. You start out with basically a horizontal and a vertical uh ground plane.

At the very least, you should be using a horizontal ground plane like this with your device under test. Um, up on an insulated wooden well, some sort of you know insulated table like this there are as I said uh standards for you know it must be X distance away from the planes and you know things like that, but at least one horizontal ground plane on the bottom. You don't necessarily have to use the vertical one. These ground planes just stop capacitive coupling to other other Uh devices and uh, things like that in the room and then you've got your listen device down here.

very short low impedance coupling because if that's a high imp pin, if you use a big long lead on there, um, then you know that isn't a very good uh grounding. uh point over the frequency range so it has to be very short, very closly connected to that ground plane device under test isolated and then you tap off to your spectrum analyzer in a proper EMC Uh facility. this would be outside of the shielded Uh chamber and things like that, but we won't go into details all right. now.

What we're going to test here: Very simply, very crude me. so don't hold me to task over this. Okay, it's just an example of pre-compliance Uh, basic pre-compliance testing. Little Uh USB Charger: One of these crappy little Usb chargers in a nonshielded box so it's going to be radiating stuff.

But of course, we're only going to be uh testing the conducted emissions of this thing. and I've got myself a ground plane. Although in this this instance, it's really not going to matter. uh, much at all.

You can do it. Basic testing without a ground plane and we've got a power supply up here. There we go. that uh, Powers our Uh 12 VTS goes into uh, the source part of our uh listen box here and we've got our coax coming out to our Spectrum analyzer there Rial DSA 815 and we'll show you how to set that up and uh, this just mounted above the ground plane here by a certain amount.

otherwise you just rip out the ground plane and put it on the bench and I've got the ground. The Source Point of you don't want to ground this side here. you want to ground the source side to your grounding plate on the bottom. So that's the setup now.

I show the Spectrum analyzer setup in a minute. but uh, as you can see, we are getting a spike here now. I've got it disconnected so it's not uh powered at all. but it is, you know, still connected uh through to the Uh device under test side of this and you'll know notice that we're getting a spike here that's a um we're going from 150 khz to 30 MHz span at the moment so that's like you know 25 MHz or something and if I disconnect this we'll find that the there we go it uh vanishes so that now with disconnected we can get ourselves a baseline.

although you can see how at the low end over here it is uh we do have some noise right down at the low end there and if we disconnect our coax from from our listen device, there we go. that's our noise floor. so you can take that noise floor as a reference and actually uh, subtract that out later if you want to. but we're not going to Bbel with that today.

All right. So here's how to crudely set up our Spectrum analyzer for this basic pre-compliance testing. It will of course depend entirely upon your Uh device under test your test setup, what version of the standard you using, etc, etc. But I'll show you a basic one uh, basic one would be 150 khz to 30 MHz frequency range.

So the first thing we want to do is go into frequency there. uh the start frequency. We want 150 khz down there and our stop frequency is 30 MHz already set up. Now we also want to go into our bandwidth detect there and our resolution bandwidth.

We want to go into that. So there it is. uh, currently set to 9 khz. That's what we want it set to for this uh, basic frequency range.

Now the filter type here. uh Gan or Emi uh. this actually this spectrum analyzer has a specific Emi filter which we want to use if we have it I think it might even be an A software optional extra, but you don't necessarily have to do that if you're using the Uh Gausian uh filter as we'll see in a minute. and if we're go to amplitude, we don't want any input attenuator at all at this stage.

so we'll leave that set to zero. DB The other really important thing is our detector type there. uh you for a very quick uh first pass measurements you want to send it to positive Peak there quasi Peak is what we're going to do for more uh, detailed measurements of this thing which uh takes a lot longer to um, scan and give you a result. but the positive Peak that's what we want with our Emi filter if you have it and what that positive Peak detection type is going to do is basically going to give you the worst case at each Uh frequency point, the worst case value and that's really all we care about.

you know, does it go over the limit or not? That's pretty much it. Now for the Sispa standard Emi Testing: the units are always going to be in DB microvolt. so we want that set up. so our vertical scale there is uh Z DB reference point is 1 microvolt.

So for example, if we'll getting 1 molt a signal was 1 molt in amplitude, then that would be 60 DB above up there. So our reference lines down there. this is our inherent noise floor of our Spectrum analyzer setup here and I think we're ready to go almost now. One thing we can do to see if we're uh passing or failing a basic limit is to set a reference level up here to show us whether or not we're past faring.

We can do that on this spectrum analyzer or Spectrum analyzer. different. Some will have it, some won't I can go into Trace pass. fail here I've already.

uh, set it up so I won't bore you with the details but I'll turn it on and basically it gives us a reference line there and the Sispa 25 standard I uh, just a generic Uh one. Won't go into details, but it's basically 1 molt uh reference level across this particular frequency range. So 1 molt um, as the example I used before is 60 DB. So uh, we want to go into the setup there and you can edit these data points and the amplitude there there it is.

Uh 60 DB is our reference level at the upper and lower frequencies so it draws a straight line. You can actually set like an envelope in there, but we just want a straight line basically. So here's our waveform and that if it goes over when we turn this thing on, if our signal goes over that purple line there. What? Warning: Will Robinson Where uh, we could be exceeding our limit so this spectrum analyzer past fail.

Things actually quite. uh, good. We can go in there, set upper and lower limits, but as you can see, I've got Um 60 DB which is the 1 Molt DB microvolts which is the 1 molt uh level. uh, crude by the standard here and we set those at both points and now we are ready to go.

We're ready to turn this sucker on. So here's our noise floor. I'll connect my coax to the listen device and still having an apply power and you can see that right at the low frequency down in there it's jumped up and you can go have a look at that. But anyway, here we go: I've got a 12volt supply here I'm going to plug in my USB charger and let's see what we get once again from 150 khz to 30 MHz with our cisper 25 standard limit of 1 1 M and plug her in Tada Look at that.

Uh, once again, we got that Peak that we uh saw before and you can see there's a bit more Broadband noise up at the high end. there? you know that 20 to 30 mahz uh region up there. but you know there's the standard so you know we're well below that look and I'll dis I'll disconnect my coax again so you can see that so you can see that's our noise floor of our system and then oh, it jumped up just a bit at the high end. But what really concerns us is this low end down here which is above that purple line.

So we're in trouble. down at the low frequency down at, you know, the hundreds of KZ to a megahertz range. so we want to zoom into that and see what's happening. So I've got my marker there and uh oh, you probably can't see it.

four. it's obscuring it. a bit bad positioning there, but you know, 4 mahz something like that. So it's up to a couple of mehz.

there's my marker point. so 2 MHz So let's go from say 0 to 5 or something like that and we'll be able to see that down there. So we'll go into frequency and our stop frequency will change that to 5 mahz. And now we can get in there and we can see now.

You can see that you know it's not. We're getting close to our limit, but the average value in there is not. you know, hugely above that line. we're close and a couple of Peaks are going over.

Now it's time. Um, because this is a very quick updating using that positive Peak detector. Now we want to go into our quasi Uh detector. So we'll go here and we'll go into our quasy Peak detector and this one will actually take some time twiddle our thumbs.

But we'll eventually get a result. And the problem there was if we go into a sweep but set it to uh 970 seconds. That's why we're going to have to wait a while and if we set it manually to 60 seconds, we can see it. Uh, start sweeping across here but this isn't going to be accurate so that is, uh, certainly going to take a while.

As you can see, we've gotten this far after. well 2 and 1/2 minutes. It's going to take about 16 minutes to do that entire sweep at the recommended Uh value of the Uh sweep speed of 970 seconds there. That one took a bit of time.

So what I've done is I've uh changed my frequency range from uh 0o htz to 1 mahz here and that's giving me a sweep Time auto Uh sweep time of 200 seconds. Much more reasonable and as you can see we we are uh with the Quasi uh Peak detector. We are going above our line there our reference line and you notice how reference line um jumps up there. That's because I set it to 150 khz I should have set it to zero and it would have gone all the way across.

But anyway, um, the uh range. As far as the standards concerned, it's only from in this particular case is 150 khz upwards. So we're even up to a megahertz here. As you can see, we're basically still over that nominal uh, cisper uh, limit.

So really, um, we probably want to take it out to a couple hundred couple of megahertz again, and uh, see where it actually falls below that. But we're definitely, uh, pushing out luck here, that's for sure with this design. And I should point out that something like this, uh, cheap, low-end uh, general purpose Spectrum analyzer isn't quite going to give exactly the same results as a proper EMC house would with their $50,000 you know, EMC uh measurement receiver. but you know we can get a reasonable indication and that's the whole idea.

that this thing allows you to do, uh, some basic in-house testings. Um, you know, cheap and easily. Basically, it's just your time inside the house, plus a, you know, a $1,500 or less analyzer. Not much at all and you can do this, uh, basic testing.

And most importantly, you can actually test things before and after you make changes to the product. So we're almost there for a 300 megahertz span using uh, quasi Peak detect here. and as you can see, it's starting to drop off there. and I can set a marker here.

where's our marker value? So anything sort of, uh, that point there? Sorry, it's a bit hard to see with the tiny little font on this thing, but um, anyway around about, you know, 2.3 let's say uh, 2.5 mehz and under, that's where we're We're probably going to be a bit concerned about that sort of stuff so we might want to look at our design and go well. what can we improve down at that low end. But as we saw with the wider frequency sweep before, any like above, that 2 1/2 MHz um it just drops off and it seems to be just fine. and there.

It is exactly the same shop, but back in the uh, real time there. So we're a bit concerned here with our pre-compliance testing down at the low frequency here. you know there's a good chance we may not pass the standard here and what can we do about it? Let's say this thing, ah, it's got to ship next week. You know we can't respin this board and add, you know, uh, ferites on the output and uh, do other stuff to reduce the Emi you know, change the uh, slew rate of signals inside, tighten up the PCB layout.

all that, uh sort of stuff. What can we do? Well, we can probably try adding on one of these um uh, Ferides, one of these external clamp type Ferites. You may have seen these on products and these are a common technique for just this thing where, well, your products finished. Oh no, but we failed our compliance testing.

We still want to ship this thing so what you can do is just add a couple of these to your uh, power wires on the output. You may have seen them and they may have been like a heat shrunk on the outside for example, added after the fact to make your product pass that EMC compliance. So let's see if it makes a difference. Here's our live display here and uh, let me clamp it over once and we probably with only one.

so we've only got like one turn in there. We really don't expect to see any different any drop at at all. In that right, not much at all. but if we wrap it around that a couple of times on just this one lead here I think we might be in luck.

So there you go. I've got a couple of turns inside that thing and bingo Look, it's dropped us down a hell of a lot. You can see that by adding an extra turn in there, we're almost under, even right down at the bottom end there. so we're doing pretty good.

Let's add another one to the negative line here. All right here we go. let's see if we can see this change live. I'll just clamp this over on my negative line there and look at that beauty.

So that's not exactly the uh, best solution to your problem here. If you can fix it inside your product, you, uh, certainly would if you had the time to, uh, you know, respin. That's why you do the in-house testing. The in-house uh, pre-compliance testing at the design stage at the Prototype stage allows you to get a quick ballark indication of whether or not you're going to pass this.

um, will pass EMC compliance whether not, you're on the border, and when you're doing Simple bench testing like this, even the ground plane isn't going to, uh, save you. really. Just be careful about external, uh, electromagnetic fields here which can, uh, you know, radiate into your test system. That's why these things are done in proper uh Shield enclosures.

Things like that. We're just mucking around here on the bench doing some basic testing as we've seen before. my LED lights above me that I use for shooting. Watch this.

If I get rid of them, we'll probably see this higher end Broadband noise disappear or lower. Look at that and I switch it. that spike is still there. Switch it on and see.

So just be careful about what things are radiating into your test system here. or capacitive or inductive coupling. Just be aware of it. And if you're wondering, does this metal shield make a big difference in this particular installation? No, not really.

Let me remove it. Live here. Oh, there we go. A bit more higher Broadband noise there, but you know, essentially it's not going to uh, change the issue that we had at the lower end.

and of course uh, radiated testing would be a whole different uh ball game again and that requires a Uh calibrated antenna which you connect up to your spectrum analyzer and a shielded anaco room I've done Uh videos on on those before those um, indoor Uh test sites for EMC compliance which I should link in down below actually. but this one was just a basic video to show you how you can do some basic Uh conducted pre-compliance measurements. and the good thing is is that you can do these measurements even if they're not. You know, absolutely correct to the absolute value.

You can see if you're getting anywhere close to the standard, uh, to the standards limit or not and if you are, then you can make changes to your product and then rerun the exact same test setup in house. Doesn't cost you much to um, spin a new design and add little you know, like you might put some little fery beads in there. you might do this or that, you might Shield this or that or do something and uh to and then you can see the changes that in your product, what they make to your Uh conducted and your radiated emissions. um as well.

So it's really, you know one of those vital things in-house EMC pre-compliance because it's much cheaper to do it in-house at the design stage than is to get a report back W Failed. So there you go. There's a very brief look at uh, some basic inhouse EMC pre-compliance testing. and don't please don't take this video as uh Gossel cuz there.

the standard is incredibly complex and it depends a hell of a lot on your device under test and all sorts of stuff. So you know and the test setups and the frequency range and the type and your product and the different classes of Uh product that you can have Within the Uh standard. So this might be you know, a class four or a class three standard or something like that, depending on what value it actually meets and the market you intend to sell into and etc. etc.

But anyway, um, there's I'm sure a lot of people with uh, uh, ton of Emi um EMC experience will no doubt point out some uh, really good links on the Forum and in the comments. so if you do have those, please, uh, add them. Hope you enjoyed it. Catch you next time! Um,.