EMI/EMC Part 1: Intro to EMI/EMC Simulation

[This was auto-generated. There may be mispellings.]

Okay, I think we'll maybe get started here. Thank you all for joining us. A couple of minutes after the hour, so we'll go ahead and get started. We've got a full hour, I think, in terms of information. Very, very complex topic today.

So I assume you guys can hear me okay, and you can see my screen okay at this point. Please let me know if you can't. I'm Mark Jones. I'm the Electronics Manager here at Ozen Engineering, and today we'll be talking about EMI and EMC simulation.

We'll be giving a demo, so I'll give a presentation for the first part of this, and then Zefan Yang, who is one of our engineers, will be giving a live demo talking about some of the aspects of EMI and EMC. So Ozen Engineering.

We're a specialist in simulation, engineering simulation, using the ANSYS tools, and that's across multiple physics, structural, thermal fluids, and low-frequency, high-frequency electromagnetics.

We are an elite channel partner for ANSYS, and we offer sales and support for the software, as well as training and mentoring and consulting services. And we have sales offices and territories on both coasts of the U.S. and headquarters in Texas.

We have a number of other companies in California, in the Silicon Valley area, as well as the Mid-Atlantic region on the East Coast. So that's a little bit more about our business.

And today we'll dive in in terms of our presentation part, and I'll begin by kind of briefly reviewing the fundamentals and requirements for EMC, EMI, and then aspects of testing and simulation, how we can use those and how they compare. And then I'll talk about PCB level simulations.

You may know that complex PCBs are a key part of EMC and EMI challenges and the work there. And then we'll talk about three-dimensional simulations moving from the PCB level to more three-dimensional structures and how simulations can be used there.

And then I'll talk through some examples in conducted and radiated emissions, and then also talk through some examples on the immunity side for electrostatic discharge and lightning strike.

So you may be aware, but in general, EMI and EMC is kind of the branch of electrical engineering that deals with unintentional generation, propagation, and reception of electromagnetic energy, which can cause undesired effects, such as interference or even physical damage in equipment.

And typically the goal of electromagnetic compatibility is to maintain good performance for different devices that may be in the same common electromagnetic environment. And typically when you start out talking about EMC, EMI, you have a diagram like this with three different components.

We have your source, which is your aggressor, maybe broadcasting noise or conducting that over a cable or maybe through the near field, such as inductive or capacitive effects. And then we have a victim device, which is receiving that energy in some way.

And so we can think of it as a source, a victim, and a coupling path. That's the three elements there. And a typical example of this you may be familiar with is an electrical drill, like a handheld drill. Those are sort of notorious for creating interference.

And so, for example, in this animation over here, you can see that it may be conducted along the power cables over to a nearby electronic component like a television, or it could be radiated through the air. And there are strict requirements on EMC.

And EMI, in general, they can be divided into emissions level and immunity level requirements. And so for emissions level requirements, it can be conducted and or radiated. And there's tests for that, both at, say, engineering level and engineering standards, as well as regulatory standards.

And then for immunity, it also covers radiated immunity and conducted immunity. And electrostatic discharge is also sort of a difficult test. That's immunity to discharge. Discharge and shock.

And then transfer impedance is also sort of an immunity where we have coupled effects, currents coupling in and producing voltages, maybe on a cable, through the shield of a cable, or on a PCB, where you have a power delivery network and you have different parts of the power plane that's coupling to each other.

And really at the heart of EMI and EMC, a lot of times is signal and power integrity concerns. And so this is sort of at the end.

So, signal integrity is very important, looking at resonances and reflections, which can cause radiated emissions or power plane resonances, for example, that can also lead to emissions. So, signal integrity and power integrity are design issues that are closely coupled to EMI.

So, motivation on the simulation side is kind of spurred on by testing. And testing is a lot of work. And testing is a lot of work. And testing in these cases can be expensive and also time consuming.

So, if we look at the cost of the test versus the time it would take to take corrective action or re-spin, you can start early with a bench level test for a given component, such as a PCB. And this is showing a bulk current injection test.

It's good to do these tests early, but you do need a physical prototype. And you do need to maybe make modifications to that prototype. Which can take some time to re-spin the board and redesign it and lay it out again.

So, this is the earliest point where people start testing and trying to make sure that at the component level, they're addressing these concerns.

And then moving into, say, a radiated emissions test in a semi-anechoic and a anechoic chamber, you're starting to then, you know, get into specialized skills, specialized hardware, maybe dedicated chamber time. That can become expensive. And at this point, a lot of the designs may be frozen.

So, you can think about having, you know, less degrees of freedom at this point because some of the systems, the subsystems may not be as easily changed at this point. So, you're getting some data, but it may be difficult to take action or more costly.

And then at the full platform level, for like a radiated immunity test for a complex vehicle, at this point, you know, there's no turning back. If you fail this type of test, it can be a little bit more expensive. It can be, you know, hundreds or thousands times more costly to correct that.

There's typically, you know, kilometers of cables in a lot of these complex vehicles, as well as, you know, tens of subsystems. So, there's a wide range of testing and reasons why we want to use simulation.

So, similarly, looking at simulations in the EMI and EMC domain, we can start looking at these at how much effort is required versus the benefits that you would get out of it. And there's a lot of reasons why we want to use simulation.

would you would start very early on with the design improvements and this could be related to root causes so thinking again in terms of rf and rf interference or filtering maybe signal integrity power integrity design this can be done fairly easily very quickly some of these simulations can run in in seconds or minutes and you can analyze you know potential issues that may cause emi and emc concerns later on so we can get a lot of bang for the buck at this at this level of testing and then moving on if we combine those different types of subsystems into a full product level simulation we're then looking at pcbs maybe they're coupled together with some drivers we have realistic sources that are needed to drive the different parts of the electronics so we're coupling electromagnetic simulation and circuit simulation together and we're looking at radiated emissions or coupled emissions from the boards and this is also very fruitful at this point but does require more care and more consideration and then lastly there's the complete virtual compliance simulation and this would be what most people would want to do this is the hardest thing to do in practice takes the largest amount of effort and largest amount of skill but at this point if you've done the design improvement you're going to be able to capitalize quite a bit at least after some of these things and then make some improvements and some of the things that he said can become very Brenden .

the good innovation for the brazilian solar system and overall the cloud technology hasihn pumps andfeld systems. There hasn't been much of an accuracy assessment by the solar system that we've done or that we have apologize because it has.

good running temperature if you want visit it once a year on a test birthday it will not verde again, a little more detail at the component level and the PCB level. We really want to focus on extracting good performance of the PCBs.

So this could be S-parameter simulations or RLCGC simulations for the PCB and the component, and then combining that with sources in the time domain for circuit level simulations for maybe the burst source or the ESD waveform or the electrically fast transient waveform.

And we can combine those together very effectively to produce radiated emissions and conducted emissions predictions from the simulation models. So PCBs are a big part of that. Another big piece of this EMI and EMC challenge is cables, cable harnesses.

And in this case, we want to control the crosstalk or the immunity or radiation from those cables. And that starts with a good cable design, twisting and shielding. And in the simulation model, we need to control the cross-talk or the immunity or radiation from those cables. We need good workflows.

We need good efficient workflows. So we have cable design kits. We have hybrid codes that can combine transmission line solutions with the full 3D solution so that we can accurately look at radiated emissions from those cables without having to create a full 3D mesh for those.

And another component that's very, very relevant to EMC and EMI is antenna design. And antennas in this context, which is a very important component of the EMI, are very important components of the EMI. And so we're looking at antennas that are designed to be able to be properly designed.

So that they don't interfere in a lot of cases with another antenna that may be on the device with a platform. And so that's looking at RF interference or RF cosite, as well as descents, which is degraded sensitivity due to a circuit that's nearby.

And so looking at antennas in the context of EMI, it can get fairly complicated. But antenna design in these simulations has been done for a long time, looking at radiation patterns and the impedance of the antenna design.

And one thing related to antennas would be electromagnetic exposure to the human body, which is also governed and regulated for different products and at different levels.

And for wireless transmitters and wireless devices, that's a maximum specific absorption ratio, or SAR level, that can be deposited as power in the body. So we can use electromagnetic simulation to measure the power of the body.

And then we can use electromagnetic simulation to measure the power of the body. That's also guided by our have potentially higher radiation time for theBlanket targets.

But also we can look at data like IDAP,窓 compartment reg unhealthy tries collect авать sigmédia for digital audio bring Надзор Buddhist the�라 embaixo cool states Iinини 맞추기 one Lastly, the full platform is obviously the most complicated part of this.

And at this point, we're looking for full systems emissions, conducted irradiated emissions and immunity.

And one example of this would be like an electric powertrain for electric vehicle, all the way from the motor and the battery pack through the cables and the cable harnesses, the inverters, DC to DC converters.

So it is possible to look at those, but it requires the combination of 3D, 2D and 1D simulations. So starting at the PCB level, which is an important part of this, it's very easy in the ANSYS tools to quickly scan the entire board for design violations that may cause an issue for EMI.

So looking at potential issues from a signal integrity and power integrity standpoint. So we have a geometry based design rule checker that's specifically intended to make sure that we do not have issues on the board.

And one example of this is a net that's crossing a split ground plane or reference return path for the current. And this can cause issues in practice.

And we can look at before we even do simulations, we can look at analytical solutions that are built into the tool to look at corrective actions, such as placing stitching capacity.

So we can look at the stitches across the capacitors across that gap, locations of the stitching capacitors, as well as the component values, and doing what if studies in real time to see if we can decrease the emissions there.

And then once we've done that type of preliminary work and diagnostics, we can run the full EM simulations using specialized PCB solvers and looking at the field levels in the near field and the far field from different types of solutions.

So we can drop in capacitors and make sure that we are meeting the design expectations. So that's a good example. And then we can look at the simulations for the PCB. So that's one example of a design issue that can be quickly addressed.

And then to do the near field and far field type of simulation, we really go from the EM side to the circuit side and back. And that's called co-simulation. So electromagnetic and circuit simulation combined together.

And the way this works is you would start off in the electromagnetic domain and extract S parameters from the PCB for the nets that are considered in the simulation. And then we would transfer that over to a schematic as a black box type of representation.

And we would connect our transient sources to that, which in this case you can see there's two differential sources there. And then once we have the solution from the transient simulation, we will push those results back into the electromagnetic model.

So those excitations are going back over as voltages and currents. And then we can compare the far field and the near fields for different types of solutions and what if scenarios. So you can see the near fields can be affected.

And then the far field levels for the spectrum can also be affected by these types of changes. And you would go through this type of loop looking at updates and corrective measures for a given design. That's combining the EM and circuit simulation.

And ultimately, we would do that to optimize the PCB design using these simulation approaches, looking at signal integrity as well as power integrity for complex PCBs. Or packages, extracting those parasitics and doing circuit simulation and looking at near field and far field radiated emissions.

In addition to that, we can look at resonant modes, which is basically an eigenmode solution or a natural resonance of the structure for the different planes on the power plane or the ground plane. And these may lead to emissions issues. They may be coupled to by different signals on the board.

And that produces radiated emissions further down the line. And we can also do an induced voltage analysis, which will couple from the far field to different nets on the board. And we can see which voltages are being excited at what frequency.

And if we have any issues that look like they would be causing an issue of concern later on. So this is another simple simulation we can do within the PCB level to really get ahead of any issues in the EMI and EMC area. So moving on to 3D simulations.

What we want to do in this case is really to create virtual test bench models for these various EMC standards. And there are challenges here because they're complex test environments. But we have hybrid solvers for these electrically large models.

And we have 3D components, which are building blocks that we can use. And then we have EMI and EMC templates for different types of problems. And so, for example, we can do a bulk current injection according to an automotive standard.

Pulling in 3D components for the transformer that's coupled into the PCB. We can do similar setups for radiated emissions, coupled emissions, conducted emissions for CISPR-25, for example. And this would include building blocks for the radiated emissions chamber and the table and the antenna.

Pulling those in very easily. Setting up the source, the listen for the conducted emissions as well as the test support structures. And really allowing you to concentrate on the definition. Of the device under test.

And we can do that for immunity as well for electrostatic discharge simulations per IEC 61000. We could do direct or indirect coupled discharge simulations. And I'll show an example of that coming up.

And then looking at the larger scale and zooming back out to the radiation hazard type of issues that may be of concern for various MIL standards. We can look at large platforms. We can look at some of the hybrid solutions that we have for electromagnetic solutions.

And we can predict where those hazardous areas around the antennas will be located. And where we need to make sure that people do not enter those locations. And then lastly, for a lot of commercial concerns, developing the SAR certification and compliance information.

This is becoming a very growing area for using simulation. In place of the measurements that can be done in chamber, which can be very expensive. So we can get a lot of good data. We have frequency dependent simulation data for the different organs of the body.

And very accurate meshing techniques that can take advantage of the geometry. So that's also another area for EMI and EMC for three-dimensional simulation. So let's look at an example of a conducted emission. An example for maybe a power electronics converter. And this comes from an ANSYS customer.

And so this is a good example showing a measurement setup for a PCB, which is a CPIC converter. And it is basically a DC to DC converter. Switching converter, which are sort of notorious for having different conducted emissions problems.

And so this example has measurement data and the test configuration as well as a model that was performed. And there's really four key steps in this type of model. The first is the passive components.

We want to accurately characterize those with their equivalent series resistance, equivalent series inductance, for example. Make sure that we have accurate frequency dependent models from those. Maybe from a measurement or impedance analyzer measurement or a data sheet.

And then the second part of that is the active components. Making sure we have very accurate advanced dynamic models. For the IGBTs or the power MOSFETs or diodes that may be dependent on the switching behavior and creating those types of conducted emissions.

So the active components are very important. And we have tools that can characterize those from the data sheets and the working point. And then the third part of that, you see the block in the middle there for the PCB. And also for the cables.

And that's creating electromagnetic models that will extract the RLGC or the S-parameters. And then the last step is the passive components. And that's the third piece there that's needed for the simulation. And we run the simulations to create a reduced order model.

Which can then be put into the last step. And that is a complete transient model that will incorporate the active and the passive components. As well as these electromagnetic models for the PCB and the cables.

And when we do this and we run the transient simulation, we can convert that over to the spectrum. And then we get, in this case, we can see good agreement. Between the simulation model as well as the measurements for the amplitudes and the frequencies for the emissions.

So looking at an example of radiated emissions, we can do a similar type of workflow. In this case, also comes to us from an ANSYS customer. Where we look at a microwave oven that was modeled a complete 3D environment with a semi-anechoic chamber and the antenna.

But also at the circuit level looking at the PCB and the sources that they're at. And then we can look at the radiation at the PCB level. And combining that to look at the radiated emissions at a certain distance away per the standard.

And if we don't want to model the whole 3D volume, we have advanced boundary conditions which can be conformal. Say around the PCB or around the device under test and around the test antenna. And these type of conformal boundary conditions can take care of a lot of the electromagnetic solution.

Without having to explicitly mesh in between those types of structures. So this can really help accelerate the runtime and alleviate some of the requirements for the computational solution.

And again in this case, courtesy of the customer, we have good agreement between the measurements and the simulation for the radiated emission spectrum. So looking at example workflow which will show how to couple the different levels of simulation. We can take the specialized PCB solutions.

In this case near fields around a PCB that's of interest. We can take that and combine that with a 1D cross section of a cable harness. And we can excite the cable harness and the PCBs and put those into a 3D simulation.

So now we're combining different levels of simulation to look at a CISPR25 standard environment. And looking at the full radiated emission spectrum in that type of test. And this is all done. Combining 2D and 1D and 3D simulation. Per this radiated emission standard.

So now looking at some immunity examples. Let's first look at this indirect electrostatic discharge simulation. Per IEC 61000. And we can do this in different ways. We can do direct coupled discharge or indirect coupled discharge from a coupling plane.

And in this case we're looking at a coupled plane which is vertically. Oriented and placed near the device under test. Which is a coffee machine. Which has some electronics in there. And maybe early prototype of a coffee machine that's made for connecting to the internet.

And other types of electronics functionality. We can solve this in the electromagnetic domain. Looking at the fields. And what field levels are coupled over to the different electronics components. And then also injecting this pulse shape. Which is common pulse shape for ESD.

With high frequency content. And low frequency content. We can then calculate the direct coupled voltages to maybe sensitive areas of the PCB. Where we have integrated circuits or chips and components. And we can do what if analysis looking at different grounding configurations.

And seeing the levels of induced voltages there for the different design considerations. And similarly this shows a direct contact on an HDMI connector. For a PCB. And in this case we can see once we have the discharge on the connector. We can see the effect on the blue trace.

Which is the received signal. By the other end of the bit stream for the board. And we can definitely take a look at issues that may be happening in digital signal transmissions. Because of these types of strikes. And then again another immunity test that's of importance.

In high interest is lightning simulation. For these cases we can use either time domain or frequency domain full wave solutions. For the electromagnetic solution. And our goal is to predict current and voltage levels at certain locations throughout the model.

And so for example this could be a small device. Such as this quadcopter that's struck by lightning. And then we look at induced voltages on low voltage controllers. So we're seeing low voltage effects. But these operate at millivolts. So hundreds of millivolts. So 50 millivolts at 300 millivolts.

It could be an important concern. And we can also look at larger scale type of simulations. Where we have a large metal tank that's struck by lightning. And then we have large piping and racks of cables that's situated nearby. And in this case we can also look at those longer cables.

And see much larger values of induced voltage on those lines. So there's quite a scale for lightning. Type of simulations. And an example workflow that I'll briefly talk about. Is basically for an indirect strike for a PCB. Where we're looking at an example PCB.

That's situated near a metal structure. That's struck by lightning. We can use 3D components in the model. Pulling in maybe predefined setups that we've already created. Or boards that we already have from a layout tool. We set up the material. And the materials and the boundaries.

As you would for a given simulation model. And then we would solve that electromagnetic model. And in this case we did a frequency domain solution. That meshes the full region. And then we get that solution. And combine that with a circuit level schematic representation.

Where we can connect those two domains. So we have a transient waveform. That is a standard current pulse for a lightning excitation. And then once we have the solution from the transient side. We can then push that back over again to the field solver. And we can look at different field levels.

That are induced around the area. Or we can look at the circuit level. And look at time domain induced voltages and currents. Say for a specific type of region of the board that may be of concern. So in summary. We've talked through quite a bit of different type of topics.

EMC EMI is a very complex subject. And there's a lot of concerns there. A broad range of engineering decisions and concerns. Things that have to be addressed. But fortunately with today's modern tools. The ANSYS tool set can address many of these requirements. And design challenges.

Looking early on especially to reduce the risk of downstream product delays. Scheduled delays. Or you know worst case compliance failures. So it's good to take advantage of those. And then the good news is. A lot of these are now integrated together. The tools are integrated together.

I've mentioned the circuit. And the EMC simulation. The 1D and 2D and 3D simulations for cables. And circuit boards. And the larger 3D environment. And as well as these workflows. That are getting better all the time. To help ease the complexity. And help with the ease of use for that.

So thank you very much for your attention. And I think at this point. I will turn it over to Zifan. And we will move on to the demo. And if you have questions. Please put those in the chat. And we will try to address those. Thank you.