Hello and thank you for joining Teledyne LeCroy today. A few housekeeping items before we begin our webinar. To keep us running on time, we will hold questions and answers at the end of the webinar, but feel free to submit your questions as they come up using the questions section of your control panel. This webinar is being recorded. A link to the recording and slides will be sent to you automatically via email within the next 24–48 hours.



Upon exiting today's webinar, there will be a short survey that pops up. Please take a moment and answer these five questions so that we can continue to provide valuable content to you. So for those of you just logging in, thank you for joining Teledyne LeCroy today for our webinar titled Debug Embedded Systems with Your Oscilloscope presented by Steve Murphy.



This is part two in our new three-part series. Today's session will focus on the fundamentals of advanced measurements and math. Just a little bit about us before we begin.



LeCroy was founded by Alabama native Walter LeCroy in 1964. Our corporate headquarters is located in Chestnut Ridge, New York, but we have sales and service offices all over the United States, as well as the rest of the world.



We started as an oscilloscope manufacturer focusing on physical layer tests and have branched into protocol analysis through several compatible acquisitions. In 2012, LeCroy was acquired by Teledyne Technologies and we were renamed Teledyne LeCroy. Our presenter today is Steve Murphy. Steve is a field applications engineer with LeCroy and based out of our corporate headquarters. Steve has held a variety of roles during his 25+ years with us, including all things sales, marketing, and application-focused.



We know there are a lot of demands on your time and we truly appreciate you spending this hour with us today. I'll now turn things over to Steve to begin.



Great, thank you very much, Hilary. I appreciate that and welcome everyone.



Welcome to our debugging embedded systems using your oscilloscope. This is a three-part series as Hilary mentioned and this is part two. In this three-part series, we're really focused on how to efficiently and accurately debug embedded systems using the scope. And the goal of the series is for you to improve your knowledge and your efficiency for both validating and troubleshooting the interactions of embedded designs that include analog, digital, serial, sensors and a large number of electronic signals.



So during this series, we covered one already but this is now part two and this part is really about measurements and math foundations, fundamentals. So we're going to look at how to locate amplitude timing errors. These are the topics that we described to be covered here. So let me just review those and then I'll go over the agenda.



So we will definitely be plotting measurement values, parameters over time. And there are a couple of tools that we can do that with, tracks and trends. So that's what one of the topics or a couple of the topics will be. We also want to use math tools that are built-in, that are made available in oscilloscopes. We'll look at serial data and triggering and decoding of that and then also mixed signals for looking at parallel data in embedded designs and how that gets integrated in with the communication and control of that embedded design.



So here is the agenda. I'm doing the intro and I'm just going to do a brief recap of part one and then we'll move into the plotting of parameters and their values through tracks and trends. And then we'll see math tools, serial, mixed signal, a summary and closing comments. And feel free to ask questions. There's that question panel in here Hilary mentioned on the right side there. Feel free to enter your questions and we'll take a look at those at the end of the webinar. We're scheduled to go about one hour here and hopefully I don't get too long-winded, which tends to happen here.



I think we're pretty good with all of this. Let's define the various terms that we'll use in this webinar. So we have an embedded system. What is that really?



What does it consist of? And let's compare that to an embedded computing system. And then thirdly, a deeply embedded system. So just very quickly, an embedded system could be something like an Internet of Things or a particularly fixed-function self-contained control system that's on a circuit board.



It has multiple passive and active devices that include analog, digital, and serial data, right? And then there is a device that's on-board to process the data and sometimes, many times that does include the memory. And then there is definitely, we need to provide power to devices and so there is power conversion on the board as well. It could be a simple thing as maybe it's your controller for your thermostat or maybe it's your washing machine or whatever it is.



The complexity could be low to high, the cost could be low to high and same with the size. So, that's generally what we mean by an embedded design that we're focused on here in this three-part series.



And so now if we add in computing capability more than what's programmed above, we're going to add in a microprocessor instead of a microcontroller for example. We're going to have memory that's going to be on the board, not resident in the device. And the amount of code is definitely increased now that we're computing. But then when we want to work with the environment, we're going to have some sensors, we're going to have something that detects things and then given that data, we're going to process it and then even control things. I think about automotive and I think about autonomous vehicles depending on the level of autonomy. And so we're going to take all the sensors around the car, whether it be cameras or LiDAR or radar and make some decisions on all of that data and then control through ECUs that control the vehicle. So that's a deeply embedded system as an example.



Okay, and then so we have a representative block diagram here that's being shown and so in the center is a processor and how it needs power, it needs analog signals like an oscillator to help run things and use as the basis for much of the timing. There are analog sensors, there's digital data, there's memory devices, there are sensors, there's serial communication through I²C or SPI. Anyway, so, and then a power management IC is there on the top to control power. So there's a lot going on in this case, a deeply embedded computing system and that cause and effect is really what we're looking at here.



The cause and effect of all those signals. Now, when you use an oscilloscope, just a quick reminder, what's going to drive what instrument you use, what oscilloscope you use. And so from an application level that will give us some insight to how many analog channels, two, four, eight, or 16 if you will, analog channels, digital channels, how much memory and digital data you want to work with? Four, eight, 16, more is fine.



There are signals within your embedded designs that have high-frequency content or maybe just low-frequency. Those will determine the bandwidth you need, the analog bandwidth that you need for your scope. It could be 100 or 200 MHz up to 8 GHz and beyond.



Of course, that's dependent on what product family you're buying. But then good old Nyquist says, hey, make sure that you're going to do at least two, 2.5 times or more sampling within each frequency. So your goal is to get effective digital sampling by at least doubling the sample rate over the frequency that you need to be measured.



And then we talked briefly about the vertical resolution and how important amplitude accuracy is and the dynamic range of oscilloscopes varies out there. Some 8-bit, 10-bit, 12-bit hardware that are out there with processing takes you beyond that dynamic range, that vertical resolution. So the amplitude and sensitivity you're looking for and accuracy you're looking for will determine your vertical resolution. Also, how wide a time window you're looking at.



If you're looking at short windows, then you really don't need a lot of memory. But if you're looking at wider time windows and want that fast sampling to maintain your effective digital bandwidth, then you need to put some memory to it. So make sure you have scopes that have enough memory so that you can have enough horizontal timing resolution through memory. Again, always look at your scope to see what the sampling rate in memory is set up to be to support your embedded design debugging.



Okay, let's move on. We also did a quick review of four different areas of tools: signal display tools, measurement tools, math and functions, and then application-specific. And we had a little display here that was a little reminder on the slide set coming up on which of these tools we were discussing.



We also have a slider bar there that shows us whether we need visual or analytical tools. I'll just say before I move on though to cover that, that today we're really about finishing up on the measurement tools and then we're going to hit math and functions and some more application-specific stuff sprinkled throughout these three webinars. So that's where we're heading. And to my point of the visual versus analytical tools, the usage model for tools that are built into scopes can be simple tools depending on how much data you have.



If you don't have a lot of data, that's simple, right? No problem seeing that. But when you take this and you look at wider time windows and there's much more data, you can't go with visual tools. You need to have analytical tools.



So that's what we do throughout this webinar is talk about analytical tools versus visual tools.



Okay, and remember the four groups of tools that I mentioned? Well, here they are. They're listed in the details of each of the signal display tools, measurement tools, math tools, analysis tools, they are all shown here. We covered all of these in our part one.



Feel free to go back and peruse and review that. Today though, we are going to focus on finishing up on some measurement tools, some other plots of measurement values and then we're really going to get into math and using some serial data and some mixed signals and putting this all together.



Okay, so that was the review. Now here we are. Let's move into plotting parameter measurement values and we'll start with tracks and move on to trends.



And what is a track? I have to put this in here. And it is a waveform composed of parameter measurements that are synchronous with the source waveform.



They're parameter measurements and the vertical units are the parameters that you're asked to track. And it could be amplitude, could be rise time, fall time, duty cycles, whichever and the horizontal units are in seconds and they are synchronized with the actual source channel.



Therefore when you look at a sourced waveform and you do a track of it, the track will have identically the same number of samples as the source waveform. And in general, the track is a tool if you want to capture a continuous stream of data, they're close together, you're making measurements but you want to find what all those values are and if there are any anomalous type parameters that are not what you expected and you can locate exactly where they are.



So let's go do that. I actually have a lab setup that I'm thinking about that I'll show you in a second but to set the stage, what is a track? And to use a track, we're going to use a math function to do that but we're going to make a track of a parameter. And what's great about parameters depending on the scope that you have is an all-instance type of oscilloscope that provides every instance, every measurement on a single acquisition and those values are recorded and added to the statistics table is important.



And so here on this statistics table of P1, which is a measurement of frequency, the last measured value here off to the right, number eight here, is over here, the last complete frequency is 955 kHz. The smallest and the largest of all eight of these frequencies is listed here.



And then of those eight, the average mean is 977 kHz. So they do vary and we can see the measurements that are placed within each period of the frequency measurements are listed there. So they are varying and they're varying around 977 kHz. In fact, they're varying around 14.4 kHz around that mean.



That's how those eight measurements are varying. So let's plot this though. Let's do this. Let's create F1 a track and the track is listed here and it's a graphical view of showing and holding the amplitude, the value for the measurement and then holding it until the next completed instance, in this case frequency, we measure that instance and then we plot that as an amplitude and we do this all the way down the line.



And each one is plotted, it's a math function and graphically you can see the data but if you put a cursor on it you're going to see that the actual value is listed on the annotation box for this particular waveform. That's how the Teledyne LeCroy oscilloscopes show you the values, if not down on the side on other waveforms. But every instance is listed and plotted and therefore we can graphically see what's going on.



See how they are time-aligned and actually the measurement and the number of samples match the channel 1 acquisition.



And there we go. But now what happens if we look at more data? Is there any variation? There is. In fact, let's do this.



I always like going to the oscilloscope. To do that, let's go back to let's do this is the webinar.



I want to go to this right here.



So this is my workbench. Let me move my can of soda out of the way.



And so what we have, this is my setup and the star of the show here today, if you will, is our embedded design.



And this embedded design has a microcontroller on it, probably an FPGA I believe is in here and we have test points that make some of these signals available to the outside world. I have probes that are connected to an oscilloscope and okay, here's the other, if you will, another star of the webinar.



And so this is an MDA, it's an 8000 2 GHz scope. But this is an 8-channel scope so I don't have to move probes around as I go through this and choose different sources to measure.



So this is the scope. This is, I got channels here, waveforms being acquired through probes.



I have mixed signals. So you can see that when we talk about mixed-signal data I actually have, I'm using eight of the 16 leads to capture mixed-signal data.



And I think that's pretty much it. Now off over here is my arbitrary waveform generator and it's generating a signal.



And I've captured it with a BNC. I am going into channel 5 of the scope and the scope is being set to 50 Ohm to match the output impedance of my generator. Now the scope is PC-based so it's running my camcorder here. Hello.



It's running my camcorder. It's also got the internet browsers and I'm connected to my Wi-Fi and so I'm connected to you through GoToWebinar. I also have the scope application running and if I didn't say it already, of course the webcam. So let's go to the scope application here and this is what it is right here and I am this is our signal.



I'm just going to shut this off and back, I'm going to go to this here real fast. So what I'm doing back over here, I'm going to acquire this particular signal into here and we have on the scope application now we can see this is the actual scope, right? Running scope application. And so let us maximize that.



Here's the signal and coming out of the generator. So we're running live right now.



Sure, you see that? Yep.



Now it's obvious here that we do have some modulation going on, right? But let's just do single-shot for the moment. Let's take P4, which is the slide set. Let's go to P4.



Let's measure frequency of this signal. So P4, we go ahead and we selected frequency right from the pull-down measurement choices and then we grabbed C5 and brought it down here. So that's our source, that's channel 5 and now we're making measurements of frequency on C5. But we talked about statistics.



So in our last webinar we covered statistics. We also did histicon here and so if we let this run normal, we can see the min, the max, and the average mean of 1,700 measurements here. We can see it's kind of bimodal. We know something's going on here.



So that is the use of having statistics in a table that gives you the range of data and then the histicon that tells you the distribution of the data. But now let's use our tracks. So to do tracks, let's go ahead. I'm going to touch P4 that gives us quick access to this track function.



And this is P4. So let's go ahead and hit P5. And now this is tracking the frequency changes. We have a new waveform, it represents the value of each of those, and we can see that the values are getting smaller and smaller, okay?



And therefore less amplitude. Now what happens if we get more data? Let's take the time base. We're right now in the bottom right by the way, 100 ns/div, 1.25 kilosamples is our memory for each of these waveforms and we're sampling at 1.25 GS/s.



This scope will go up to 10 GS/s and other scopes go way beyond that. But nonetheless, let's take this and slow this down and let's go get another single-shot. Oh, we noticed that as we grab more data, that we do have some changes in frequency. It's modulating and we can see now that it's increasing, decreases, increases in frequency.



Let's go get some more data. You'll get the idea here that we are now doing a frequency modulation which is now apparent with our track.



We're going to use the track for some other things here in just a moment but anyway, we can now see how that frequency is changing. In fact, I'm going to grab P4 over here, bring it over to P5, it duplicates it. And then I'm going to go ahead and say, hey, F5, that's my source of frequency. So now we can see that our frequency modulation is actually 100 kHz. And so that's the details here on this particular waveform so that we can see the actual frequencies and how it's changing. And then when we see what that rate of change is and we measure that, it's listed there as 100 kHz. So when we talk about visual versus analytical tools, we need these analytical tools when we have this much time window displayed when we're trying to analyze the interaction of an embedded design with measurements like this.



Okay.



And I am going to do this, I'm going to go to Window. I want to move this up here and I'm going to put this down here so I can cycle between my Windows applications easily.



Let's go to this and same thing here. Might as well do it now before I... Okay.



Yep, so I'm liking that. And then, so this is, I'm going through the slides right now on PowerPoint running on this PC and this is, I'm taking you through this, right? So that is a track and I want to go somewhere else. Let's go back to the oscilloscope application.



Let's hit this again with not from the frequency modulated arbitrary waveform generator signal that I'm simulating from my embedded design. What I'd like to do, let's go up to file and I'm going to recall a LabNotebook. I'll describe what that is here and I want to look at this track. Let's flash this back.



And while it's doing that, what it's doing is changing the whole scope, the panel setup. It's taken those waveforms I had previously and throwing them away. It's bringing in new waveforms and this is what we have here. And just to, again, point out what a track is, the signal source now by the way, is coming from here and so we have asked that output to give us the signal.



And let's, I guess we can just... well, I'll just keep going the way I have it.



And if we come back to this power, I'm sorry, to the scope, we can see that we have three parameters turned on. P4 is the width of C1 and C1 is right here. We are sampling at 100 MS/s at 1 millisecond, so 10 milliseconds of time with 1 million data points on this waveform.



We then turned on width and we said, okay, give us a track.



And the way we did that is we said track of F1 using F1 of P4 and we are plotting on this waveform and you can see what these pulses are.



I'm just going to do a zoom here.



This is a zoom and we can see pulses are wider and wider and then it drops down. And that's really showing right here wider, wider, and then it drops down. I could do a zoom and tie it in with this as well so you could see that track.



But indeed, this is showing you every value plotted, held, and in this case, and showing width.



And we can see that it is a repeating pattern. And we didn't know that just by looking at the waveform or making measurements on it or using cursors. We use a track to help us understand that.



Okay, now another little signal source I have is off over here. This is a Microchip PIC. It's an 8-bit PIC that I'm powering from USB over here. And I've got a probe here.



This is a probe coming in off of, we call it channel 2 here, that's going into the channel 4. And so let's do this. I'm going to go back to the scope setup and just quickly allow me to go to the file, recall a LabNotebook and let's pick these measure tools. This is going to be pulse-width modulated and it's going to just show us the track but I also will be able to show a trend.



And I'll mention that in just a second. So now we're flashing this back.



Our source is that little Microchip PIC which generates all these different signals and here we go and let's just go live.



Yes, so live. So we're triggered here in the middle. This is pulse-width modulated. We are edge triggered. So it's just giving us any update and we can see there's our edge and it is pulse-width modulated, But let's go ahead now and select P7.



Let's go to value here. Let's click here and turn on statistics. You can see the first tab, by the way, for those of you new to Teledyne LeCroy. The first tab tells you what dialog menu you have open.



It's the parameter one. Let's turn on statistics and histicon. What is this telling us? It's telling us that indeed we have the average mean is 197 ?s but we have a min that's in the nanosecond region and the largest is still in the 241 ?s area.



But of these 12,000 measurements, it's worth around 197 ?s but this is an outlier and when you look at this menu down here, you can see that we definitely have an outlier.



Okay, so let's go ahead and turn on track. I'm just going to touch this. Let's go to track and say, okay, track is set up for F1, we'll use that. This is our track and let's go get more data here. I'm just touching C4, let's go get more data and I'm going to do a single-shot.



These are all of the values of P7 called width and each value, if you turn on a cursor and go just do an absolute cursor, you can see each value is listed down here as small, which we call 157.6 ?s and then here's... I don't want to do that. I wanted to grab the mouse, the cursor, there it goes. And up here, boy, all of these are about the same. Interesting, this is 240.



So we can see this is the range of value. Let's do another single.



Watch what happens. There is a glitch that's occurring, right? That shows us in the table. Oh, there's not just one, there are two.



Look at this, right? So this is a track. It is a plot of values of width in this case and when we go ahead and take a look at this in here, this is telling us, time-aligned, that we have a problem, Houston. We call it here in the United States.



And this is Z4, let's do that. So we have actually a glitch that's occurring right in here. I just did a zoom just of that. So that's what the problem is.



And there's another one over here. So you can use the track time-correlated on any of your waveforms in your embedded design to correlate these issues, these anomalies with other parts of your design.



Okay. Whether you're using mixed-signal, whether you're using serial data, you're using multiple other channels, all of those, you can use these tracks to help you identify anomalies.



Okay, that is this.



Let me just jump to now. So that's another example of a track of a parameter. It plots values. Let's go to the PowerPoint.



I showed you this but now we want to move on to another measurement tool, which is a trend. And very briefly here, what is a trend? It's a plot also of a waveform that's composed of parameter measurements arranged in order that they were made. The vertical units are the source of the parameter, which is voltage or time, rise time, fall time, widths, whatever we were doing before.



Now we can plot those as amplitude values and then the timing is no longer the time per div but the measurement number. So it'll just keep making measurements.



If there's a unique number, it will just keep one of them and our display is now in measurements per division. How is that useful? Well, since it contains a measurement parameter, it plots it. We could actually use this as a strip chart recorder to look at things like power line monitoring, output regulation, ripple, whatever. And so this looks like a strip chart recorder can be used on multiple acquisitions and plot the values versus time. So let's go back to the scope and now I'm going to say, hey, P7, I'm picking on P7 here today.



Let's go ahead and trend this. And I want to compare a trend versus a track. Let's use F2.



Okay, and there, what do we see off to the right?



We see here the trend if we look down, we're at 100 measurements per division and our vertical scale is 50 ?s/div. Why? Because we are plotting the values, the measurements of each of these with values and there's the signal. So if I go, now we captured and I'm not triggered on it.



It just happened to be an anomaly that was captured. In fact, there are two of them. Let's go do another acquisition. Nothing there.



What happens? All those measurements per division, a hundred of them are moved on. Here's another acquisition, no glitches, nothing there, nothing's going on.



And so what we're seeing ah, there is.



What we're seeing down here, yeah, the previous, I missed it. I already acquired other data here but there is another glitch that occurs. So these are about one, two, four, six, about eight captures and this is the accumulation of all of those values. These can be exported by the way or these values can be exported in Excel or comma-separated variable and analyzed or you could do pass-fail testing on this stuff too as we covered on part one.



Okay, so this is a trend and a track and as a comparison and I'm just going to show you one more thing here.



As you look at tools, don't forget our histogram, right? This is our F3 histogram. This told us this is we showed you a histicon, but this is the... And if I just let this run, of course, now all of the measurement tools are being utilized here in this capture. So when you're debugging your design or validating it, don't forget, make the measurements as many as you want, but you can also track, trend and histogram the results. So I hope you find that helpful in helping you to have some tools at your fingertip that you can quickly use to debug.



Okay, and I took you through this. All right, now let's move on to, I'm just doing a time check for myself here.



Okay. Good. Let's do math.



Okay. And so this is where we are in the agenda now.



Math tools. This is where we are in the usage model. We're going to do some basic math. I'm going to actually extend into analytical a little bit but let's start off. Let's say we have an embedded design, we have serial data and we're sending out plus and minus.



It could be CAN bus, it could be USB, it could be all kinds of serial data but we want to use the scope's ability to basically subtract one from the other just like a transmitter sends and a receiver receives, it sends it out and receives it differentially. So let's do that. This would be our plus and our minus signal on a serial data. The result is this math function channel 1 minus channel 2.



Again, this is basic math, but let me just quickly review it. Here are the choices, some of the choices for basic math, the average, difference we're using here, invert, product, take voltage times current. I'll show that to you or the ratio can be done here too, sum. And so that's what we're doing.



And if we look analytically, look at more data, if we do a zoom of the subtraction, we can see the result can be this basic math function.



And yeah, so that's the use of this in your design. I threw this in here. This is what an example of basic math is: voltage times current. So on the top grid, the voltage on channel 2 is current, is power factor-corrected as we're pulling current in phase with voltage. If we multiply those two together, another basic math function, we get this waveform. We can also look at the output voltage and the output current as this blue and green waveforms, multiply those two and we get the output power, voltage times power, right? And then we can then use our parameters like we were doing to measure things like input power, output power, and even do parameter math, which that ratio here would give us efficiency.



So that's the use of basic math. Let me just check this one question here. Yes, yes.



Actually I have a question here. I'm just going to go ahead and answer it. And when doing math like on differential signals, how closely matched are channels for time sync? Yes, if we're using the same probes, the scope is calibrated on the output and the same probes will give you the same propagation delay but use a source, a deskew calibrated source and verify that both waveforms using both probes line up perfectly. If not, you deskew them and make sure those differential signals, very good question, that those differential signals do propagate and arrive at the oscilloscopes' BNCs simultaneously. Fine adjust them out if they're not and then now your differential measurement will be more accurate. Yes, thank you very much for your question.



Okay, there are other math tools. This is just around functions. You can see there are all kinds of choices here.



I like to use deskew which is just perfect. I should have showed that one as a response to your question, but there it is. This is one of the tools you use. It's also in the channel menu on your oscilloscopes, at least on the Teledyne LeCroy oscilloscopes it's located up front.



It also can be a math function. Another one I like to point out is integral so we could take power, watts and integrate it to get joules, which is watt-seconds, energy over time. And then I like rescale to look at things like changing to temperature and changing to torque, changing to... using differential probes across a current sensing resistor. There is a, why don't we grab voltage signals, put in the value of the resistor, rescale it so that we get Amps per division amplitude.



So those are really good. And then zooms of course. So this is the presenter's common functions.



I also, one math function that I'm showing off a little bit, sorry, but don't forget about FFTs. In your math functions, in your math choices, there is a grouping called frequency. And so you could go ahead and use your zooms on a signal, show the zooms and then you use the math function as an FFT and you can vary the spectral content and understand the frequency content. So when you're looking at different parts of your embedded design, you can compare, hey, am I getting any crosstalk? Am I getting conducted emissions?



And so these are tools to help you with your embedded design.



Okay, moving right along.



This next section is about physical layer, electrical serial tools. And so that's what we are looking at here on this next agenda item.



Okay, And let's focus on trigger and decode although I am going to set the stage a little bit. When we look at our embedded design here and this is our little representative one, right? And it has everything in it. I can take some of these to make my point here but there is, when we look at our processor, it is talking, it is getting its control, it's getting data from... it's actually mixed-signal also data there is, I would like to show... I'm bouncing around on you, but I'd like to show digital data that's being addressed and accessed and written to and all that.



But we also are looking at serial data. We're pointing out here I²C, SPI, and I²S. It could be so many other things. So typically, here we have on-board low-speed serial data, I²C, SPI, QSPI, all kinds of things that could happen.



And we also are wanting to see the interaction of off-board data that comes in or goes out like 1000BASE-T or 1000BASE-T1, right? Or CAN or CAN FD. So there's all kinds of off-board stuff that we can use. And by the way, we can get these signals coming into the oscilloscope channel but feel free to use your oscilloscope's digital lead set so that you can acquire the clock and the data example of the I²C power management IC communication and capture that through your digital lead set. That way you can see what the commands are and then you can probe these directly which are DC-DC buck or boost converters here.



So this is, again, this is what we're pointing to as tools to debug embedded design. So let's go apply and let's look at the oscilloscope now, what tools can be available. Here I'm going to be an overview here on these two slides coming up. This one, which is triggering.



We want to trigger on particular protocol data or addressing or controls. And we want to then display the physical layer waveform in a way that makes use of it. So some color coding would be nice or put it in a table and have some interaction with viewing the zooms of that with other signals that we've acquired simultaneously. And then what we'll do is we've been talking about measurement tools.



Why don't we use measurement tools here with serial data and even extract some data and graph data that's embedded in serial data so we can say, ah, that power management IC when we address that and state this address in particular data then yes, let's go ahead and capture it. And of course in the serial world, so important to understand the signal integrity.



And so on part one we covered eye diagrams and doing them with both persistence displays but also using slicing, I should say, serial data. So anyway, these are, this is a high level serial data level of PHY tools. And on the trigger, there are lots of triggering choices that oscilloscope manufacturers provide you. We program our trigger FPGAs so that they can address the different protocols.



Why? Because I²C or UART or CAN, CAN FD, FlexRay, 1553, on and on, right? They all have their unique aspect that we can use as a reference point for capturing pre-trigger and post-trigger data. And so that's important for a trigger.



And so there is a conditional setup based on the protocol that you're using. You can see there's an example of it. In fact, let's go back to our display here. I'm going to go and use a little utility that's going to control the output of this and let's go ahead and make this be I²C.



Right? So I'm going to go to the little utility, I'm just going to go I²C, I'm going to load the demo which is communicating to the oscilloscope. Let's go back to the oscilloscope now and it's making the change and we're live right now. So we are using channel 1 and channel 2.



So we have clock and data coming into channel 1 and 2.



We're live. We control this by the way through USB so we're making this board do specific outputs that we want. And if we go back to the scope, let's just talk about tools for serial data. Now what we're looking at is a, this particular grid is a three-grid display.



I like it. We're looking at the data and the clock. Sometimes you want to go ahead and overlay those right. Let's just do a single-shot and let's just go ahead.



I'm going to, on the edge here, I'm going to do... I just want to do a zoom here.



Not too much, Steve?



So we want to look at the timing, right? So this is the physical layer, the electrical signals, the magenta is the clock and the data is in yellow. And so we want to look at that timing many times with I²C and that's why you'll use your oscilloscopes to do that. Then when we have enough data so we see the full data frame, the whole signal, we can then go ahead and annotate it in color.



And then we are doing the zoom over here. So let's take channel 2 clock and bring it back down here and we can see this is the zoom. We can move that with our mouse over and we want to validate, we want to confirm that we're happy. Let's go push this in and go a little faster.



So these are the component parts of a serial data. And so we have a start bit for this particular I²C. We have an address and we have read or write, so the direction, which way is it going. There's an ACK bit that's put in there by another controller.



And the whole frame by the way is highlighted at this particular color. Control data highlighted as a different color within each different protocol. And then of course common amongst them is the actual data is highlighted in blue, annotated in blue and there's the ACK bit in its color as well. So you get the idea above: we're displaying the data.



I meant to, let's see this. You can view this in binary. Yikes! There's a table down here that you can expand to view the ones and zeros.



Some folks still like it, but let's go to hex. And then of course let's go to ASCII. You can see how we had this thing programmed, right? But nonetheless, there is the data and all you gotta do is just touch one of these and if you want to just see rewrite addressing.



So there is, when you go ahead with serial data with your oscilloscope, you can take this protocol stuff, you can look at the direction, define the direction if you want, read or write equals.



Let's just go write, hit okay and enable it. So I put it in the wrong spot. But anyway, I put it in the address. So you got the idea. Clear that one.



But you can define your length, you can define the direction that you want. Do I have it here now? Let's try it one more time. Okay, and enable it.



No, I am looking for an address. I could just go with the Enable column for read-write. There you go. Much better, Steve.



And enable it.



Read.



Let's go ahead with R, okay, and enable it.



Okay.



So the reason I had two filters on, so I'd ask for a specific address which I incorrectly set and then, but I'm going to move on. Let's see. Oh no, I'm just now committed.



Okay. Here's just the direction of reads. There's address 31 and 30 that are right here at 3F4143. This is the address length in bytes. There's all the individual data. Okay, Steve, you did it. All right.



So anyway, there's that plus under Zooms by the way, just you can do advanced searches. You can search for particular IDs and stuff. So when you're trying to isolate problems in your embedded design, use your trigger, your decode, your search tools, and let's just touch trigger real fast. This is I²C.



You can see our trigger dialog. I went to serial. This is I²C. This is the unique protocol for I²C where our data and clock have already been set up.



I did that automatically but we could put in specific addresses here like 42 and trigger on that.



And anyway, due to time, I'm going to move along but our address that is in particular address 7 or 10 bits in this case and in a range, out of range equal to a specific value, you can put that in here and then you can say, "Hey, I'm looking for that specific byte number 12 and for one bit or one byte length or multiple byte lengths" and here's the value. Anyway, that's the idea. This will really help you fine-tune working with your serial data in your embedded design to correlate what's going on. Okay.



And then, so that's the trigger, the T.



I've been showing you the D. Let's go back to PowerPoint to stay on track. That's the T, that's the D, that's the D I took you through and I spent some time on the table. So I've already covered all this.



Feel free to use on embedded designs. There are many protocols, several, and here in this case, I'm showing you the ability to use four protocols. So if you want to look at the interaction of time, scopes are awesome time scopes. Don't forget to use your scope to do timing from one serial data to another or from an analog signal to another.



More on that in our part three, by the way, but nonetheless, use these tools to see the gateway timing between serial communication within Embedded Designs. Okay.



I'm going back to PowerPoint and I'm going to do this next time. I'm going to extract on part three data from specific embedded designs, specific controller. In this case, I think it's angle of rotation of a steering wheel that was embedded in a CAN frame. Is it CAN?



No. CAN FD value. We use the parameters to plot for every particular serial data frame with a specific address in a particular location of the data field to extract the data, plot it and do this as a track. Remember that track?



There you go. Here's a DigRF signal we can do with CAN, I²C, SPI, many of the protocols are all able to be measured, data extracted, plotted and compared to other signals that could be embedded, if you will, de-embedded signals.



So at any rate, that is that. Let me just check out this.



I will save that question for the end.



Okay, and then let's move this back.



Okay, moving right along. We have mixed-signal digital capture and integration. So let's look at that. Back to our embedded design here as I had mentioned, being able to see what's going on with memory, seeing now that what the, through the mixed signal we can get things like I²C to see what that power management IC is so we can use our other channels to measure the actual voltage rails as an example of the buck or boost converters.



We are looking at actuators and sensors and other stuff that's all coming in here.



And so let's go ahead. Let's just go live on this once again. I am now going to work with our embedded design again. I do have a, let's program this to, hang on one second.



Let's get some mixed-signal going on here. So to do that, let's just do this in load and now I'm telling my device to go ahead and output digital data. Let's go to the scope.



There we go. And I am actually, if we look at the trigger, we're normal, we're looking for a pattern and this is the digital lead set connected to this part of the board and we see, I put on a cursor here but this is data 0 through data 7. So I have eight digital channels that I'm acquiring and you can say pretty much it's a counter and I've asked to view not just eight, not 16 of these but just eight of them. And I've asked for the trigger condition as well to be on this.



I'm going to just touch trigger as I move quickly for a particular setup pattern. I want it to trigger D4 on 1, D3 on 0. You can see the rest of them here. You have choices on whether they are a 1, a 0, a falling, a rising, or you don't care.



So, and then I have bus view turned on so that quickly we can see that's a 1 and a 0, a 1 and all the way through. Actually, this is most significant down here to least significant and we could, we can see the measurement right here at this particular location is listed here as hexadecimal D6 but we can also see it, the bus right here is providing the view of it.



So I'm displaying it. I'm displaying the group of data. I'm also set up the trigger to trigger on that. Our trigger location I am looking for is right here and that was that pattern that I showed you on the trigger.



So again, this is all integrated. If I turn on, for example, channel 2, we can now see that we're running at channel 2 at about 3.3 V. That's being output. It's just an example of another control signal or something that I want to use as a reference point with my other channels of the oscilloscope.



And so we want to see what that timing is in relationship to these signals.



The interaction is the word I was looking for. Okay, so this is the interaction for digital signals and for closing comments now as we are right on track for time.



In summary, we see that embedded designs take a lot, there's a wide range of signals and the oscilloscopes are really good to see the analog, digital, serial, sensor data all brought in. We focused today on additional measurement tools, math tools. I just did a little bit of intro on this stuff.



So much more to cover but I hope this was helpful. We also looked at serial trigger decode. We looked at mixed signals and how to use all of that in validating, debugging all the different signals that are available that are in an embedded design. But next week, now that we've covered all of these, not next week but it is actually, but the next part three, better put, we will apply all of that and we're going to be focused on specific design applications.



I have at least four of them that I will be working with.



Feel free to look at part one back on our on demand but there's also a whole bunch of other webinars that have been recorded that are available for you to view.



There is part one but next part three, it will be coming up. So feel free to register and join me, join us for that, which is practical real-world examples of embedded system validation and debug. Hey, I also like everything that you saw me working with on the oscilloscope that application software is available to you. You don't have to own a scope. You can just download it, install it on your PC and get a license from us after you install it. And then all these different applications for embedded designs, you should even expand that to that list, can be made available to you. And if you want, I've got LabNotebooks here that I've used, you saw me recall that, that also can be available as well.



So all the data, all the waveform data and panel setup is going to be recalled and worked with on your own.



Okay.



See, and then, so really thank you for your time, your interest. I hope you found this helpful. I want to go back to the questions here.



We're still good on time, Hilary?



Still there?



Yeah, I mean, we're at three so people are welcome to stay and listen to the questions.



But I would like to thank Steve for his time. I'd like to thank all of you for attending. And like Steve said, please feel free to go back and watch part one. We hope you'll join us next week for part three and definitely check out our events page. We are posting new topics all of the time and we're open for suggestions. If you do watch this later on demand and think of something new to ask, go ahead and send either myself or Steve a question. His email is in the slide deck.



So do you want to answer that last question?



Yes, quickly it says, could you quickly go over where the DC/AC coupling feature for the scopes might be enabled or disabled? For sure, let's just go right to the oscilloscope and I'm touching channel 2 descriptor which means that the dialog menu will open for and that's where you would access it right here. This is all the vertical controls but it does include the coupling where you would select AC or DC coupling for 1 MOhm. And as you know, of course with 50 Ohm there is no AC coupling available in oscilloscopes. You would have to add your own devices but nonetheless, so that's where it is and that's where you enable or disable it.



Great, appreciate your question.



All right, everyone.



Thanks so much. Have a great afternoon. We appreciate you joining us today. Bye bye.



Yes, bye bye.