WEBVTT NOTE This file was exported by MacCaption version 7.0.13 to comply with the WebVTT specification dated March 27, 2017. 00:00:11.261 --> 00:00:12.930 align:center line:-1 position:50% size:38% At the end of the last lesson, 00:00:12.930 --> 00:00:17.559 align:center line:-1 position:50% size:58% I mentioned we would be looking at some problem-type signals in this lesson 00:00:17.559 --> 00:00:21.688 align:center line:-1 position:50% size:45% and then use advanced triggering to isolate them. 00:00:21.688 --> 00:00:24.942 align:center line:-1 position:50% size:51% What do I mean by "problem signals"? 00:00:24.942 --> 00:00:30.948 align:center line:-1 position:50% size:58% One of the main reasons engineers use oscilloscopes is to debug their designs. 00:00:30.948 --> 00:00:34.076 align:center line:-1 position:50% size:46% That means that when they're testing their designs, 00:00:34.076 --> 00:00:37.287 align:center line:-1 position:50% size:54% they're often looking for the unexpected. 00:00:37.287 --> 00:00:40.123 align:center line:-1 position:50% size:40% They're looking for the needle- in-a-haystack event 00:00:40.123 --> 00:00:44.378 align:center line:-1 position:50% size:39% that might cause their design to unexpectedly fail 00:00:44.378 --> 00:00:47.631 align:center line:-1 position:50% size:58% or perhaps not meet desired specifications. 00:00:47.631 --> 00:00:53.804 align:center line:-1 position:50% size:57% Hi, I'm Johnnie Hancock, Product Manager for Keysight Infiniivision oscilloscopes. 00:00:53.804 --> 00:00:58.475 align:center line:-1 position:50% size:53% Let's now look at some problem signals and see how we can isolate them 00:00:58.475 --> 00:01:01.645 align:center line:-1 position:50% size:49% with advanced parametric triggering. 00:01:01.645 --> 00:01:05.232 align:center line:-1 position:50% size:42% Now, I'm probing a clock signal, basically a square wave. 00:01:05.232 --> 00:01:08.360 align:center line:-1 position:50% size:44% This happens to be one of the built-in training signals 00:01:08.360 --> 00:01:15.409 align:center line:-1 position:50% size:55% of this particular oscilloscope that we use for training-education-type environments. 00:01:15.409 --> 00:01:21.665 align:center line:-1 position:50% size:53% Everything looks normal right now, but let's zoom in and take a closer look. 00:01:21.665 --> 00:01:24.167 align:center line:-1 position:50% size:56% There's one period of the clock on screen. 00:01:24.167 --> 00:01:27.921 align:center line:-1 position:50% size:47% If you look closely, you might see something flickering. 00:01:27.921 --> 00:01:33.552 align:center line:-1 position:50% size:57% I'm going to turn the waveform intensity up so we can see it more clearly, 00:01:33.552 --> 00:01:35.971 align:center line:-1 position:50% size:44% and there we can see something popping in there. 00:01:35.971 --> 00:01:39.099 align:center line:-1 position:50% size:50% Sometimes I call it a ghost waveform. 00:01:39.099 --> 00:01:43.603 align:center line:-1 position:50% size:36% It's a very infrequent glitch. 00:01:43.603 --> 00:01:45.314 align:center line:-1 position:50% size:31% This scope uncovers it. 00:01:45.314 --> 00:01:47.190 align:center line:-1 position:50% size:45% Most scopes won't show you this. 00:01:47.190 --> 00:01:51.278 align:center line:-1 position:50% size:44% This scope has an extremely fast waveform update rate. 00:01:51.278 --> 00:01:58.076 align:center line:-1 position:50% size:56% Up to 200,000 times per second, it can update and retrigger and show you, 00:01:58.076 --> 00:02:04.791 align:center line:-1 position:50% size:58% and it can uncover the needle-in-a-haystack, problem-type signal like this. 00:02:04.791 --> 00:02:06.418 align:center line:-1 position:50% size:34% How can we isolate this? 00:02:06.418 --> 00:02:11.590 align:center line:-1 position:50% size:44% If we can isolate it, then perhaps we can find the root cause, 00:02:11.590 --> 00:02:13.675 align:center line:-1 position:50% size:37% what's causing it to happen. 00:02:13.675 --> 00:02:20.932 align:center line:-1 position:50% size:58% The way we isolate things like this is to uniquely trigger on this occasional glitch. 00:02:20.932 --> 00:02:26.229 align:center line:-1 position:50% size:58% What's unique and different about the glitch versus the normal clock signal? 00:02:26.229 --> 00:02:30.609 align:center line:-1 position:50% size:40% What's unique and different is it's a very narrow pulse. 00:02:30.609 --> 00:02:36.573 align:center line:-1 position:50% size:59% I can see the pulse width of the clock signal, I'm at 2 µs per division, 00:02:36.573 --> 00:02:42.871 align:center line:-1 position:50% size:49% 5 divisions, it's about 10 µs, a normal positive pulse on the clock. 00:02:42.871 --> 00:02:49.503 align:center line:-1 position:50% size:53% I can see when this thing pops in there, it looks like it's about 1 µs wide. 00:02:49.503 --> 00:02:53.298 align:center line:-1 position:50% size:40% Let's go into the trigger menu and see what our options are. 00:02:53.298 --> 00:02:58.512 align:center line:-1 position:50% size:53% I'm going to press Trigger, Trigger Type, 00:02:58.512 --> 00:03:01.473 align:center line:-1 position:50% size:60% and you can see we're triggering on an edge. 00:03:01.473 --> 00:03:06.728 align:center line:-1 position:50% size:46% Our other options are Pulse Width, Pattern, Rise/Fall time, 00:03:06.728 --> 00:03:11.233 align:center line:-1 position:50% size:45% Set Up and Hold, Video, and Serial. 00:03:11.233 --> 00:03:15.070 align:center line:-1 position:50% size:53% The one that's going to work for us, and we'll get to some of the others later, 00:03:15.070 --> 00:03:21.952 align:center line:-1 position:50% size:55% is pulse width triggering, and remember, it's a narrower pulse than a normal pulse. 00:03:21.952 --> 00:03:26.206 align:center line:-1 position:50% size:42% Here it says Source Channel 1, that's the input I'm using. 00:03:26.206 --> 00:03:31.128 align:center line:-1 position:50% size:44% I can say, is it a positive pulse or a positive glitch or a negative? 00:03:31.128 --> 00:03:35.841 align:center line:-1 position:50% size:44% It looks like a positive one to me, so that little icon means positive. 00:03:35.841 --> 00:03:39.136 align:center line:-1 position:50% size:47% Then, I can put a time qualifier on it. 00:03:39.136 --> 00:03:44.683 align:center line:-1 position:50% size:56% Less than, greater than, or within a range, and then the specific time. 00:03:44.683 --> 00:03:49.396 align:center line:-1 position:50% size:45% Right now, it says a positive pulse for less than 30 ns, 00:03:49.396 --> 00:03:51.690 align:center line:-1 position:50% size:53% so right now, the scope is not triggering. 00:03:51.690 --> 00:03:54.234 align:center line:-1 position:50% size:45% If I clear screen, you see nothing, 00:03:54.234 --> 00:03:58.113 align:center line:-1 position:50% size:41% because it's looking for a pulse narrower than 30 ns. 00:03:58.113 --> 00:04:06.872 align:center line:-1 position:50% size:54% Let's start increasing the value up to somewhere around 1.5 µs or 2 µs, 00:04:06.872 --> 00:04:14.087 align:center line:-1 position:50% size:38% and it should start triggering on that narrow pulse. 00:04:14.087 --> 00:04:16.882 align:center line:-1 position:50% size:23% I'm up to 500 ns. 00:04:16.882 --> 00:04:19.634 align:center line:-1 position:50% size:15% There it is. 00:04:19.634 --> 00:04:24.264 align:center line:-1 position:50% size:61% Now, it's triggering about one time per second. 00:04:24.264 --> 00:04:27.142 align:center line:-1 position:50% size:43% That's how often it is happening. 00:04:27.142 --> 00:04:32.689 align:center line:-1 position:50% size:47% If I change the time base here, you can see that it's a narrow glitch 00:04:32.689 --> 00:04:37.944 align:center line:-1 position:50% size:60% buried amongst thousands of normal pulses, we've uncovered it. 00:04:37.944 --> 00:04:43.408 align:center line:-1 position:50% size:54% At this point, what's causing that, I could take other channels of the scope 00:04:43.408 --> 00:04:47.746 align:center line:-1 position:50% size:50% and probe other signals in the system and see what's correlated to it, 00:04:47.746 --> 00:04:50.749 align:center line:-1 position:50% size:57% and that might help me find the root cause. 00:04:50.749 --> 00:04:52.042 align:center line:-1 position:50% size:33% Here's another example. 00:04:52.042 --> 00:04:53.919 align:center line:-1 position:50% size:40% This is a fairly complex signal. 00:04:53.919 --> 00:04:58.340 align:center line:-1 position:50% size:38% It's a series of digital pulses, they're somewhat random. 00:04:58.340 --> 00:05:00.967 align:center line:-1 position:50% size:51% You might think of them as serial data, 00:05:00.967 --> 00:05:03.887 align:center line:-1 position:50% size:40% and we'll get to that, I believe, in the next lesson. 00:05:03.887 --> 00:05:06.765 align:center line:-1 position:50% size:59% You can see different pulse widths, let's run, 00:05:06.765 --> 00:05:09.935 align:center line:-1 position:50% size:48% but the signal does have a problem, signal integrity problem. 00:05:09.935 --> 00:05:15.565 align:center line:-1 position:50% size:48% I'm going to zoom in, and if you take a close look at this rising edge, 00:05:15.565 --> 00:05:19.236 align:center line:-1 position:50% size:45% I'm just triggering on rising edges, 00:05:19.236 --> 00:05:21.863 align:center line:-1 position:50% size:64% again, we can see something flickering up here. 00:05:21.863 --> 00:05:28.662 align:center line:-1 position:50% size:50% Let's increase the waveform intensity. 00:05:28.662 --> 00:05:32.374 align:center line:-1 position:50% size:36% Normally, the edge has a nice smooth rising edge, 00:05:32.374 --> 00:05:36.545 align:center line:-1 position:50% size:37% but sometimes it goes up, it stalls, and goes up again. 00:05:36.545 --> 00:05:42.551 align:center line:-1 position:50% size:49% This is called a non-monotonic edge, "mono" means "one," one edge. 00:05:42.551 --> 00:05:46.429 align:center line:-1 position:50% size:57% It's got two edges for a complete transition. 00:05:46.429 --> 00:05:50.767 align:center line:-1 position:50% size:37% Let's measure the rise time. 00:05:50.767 --> 00:05:54.604 align:center line:-1 position:50% size:38% A normal rise time not on the non-monotonic edge, 00:05:54.604 --> 00:06:01.486 align:center line:-1 position:50% size:49% on the one with the monotonic edge, measures about 120 ns. 00:06:01.486 --> 00:06:04.155 align:center line:-1 position:50% size:47% To isolate the non-monotonic edge, 00:06:04.155 --> 00:06:10.078 align:center line:-1 position:50% size:50% let's go back into the triggering menu, select Trigger Type. 00:06:10.078 --> 00:06:15.041 align:center line:-1 position:50% size:50% Right now we're triggering on edge, last time we triggered on pulse width. 00:06:15.041 --> 00:06:19.129 align:center line:-1 position:50% size:37% The right selection this time is Rise and Fall Time. 00:06:19.129 --> 00:06:23.592 align:center line:-1 position:50% size:44% We can set up a qualification to trigger on the slower rise time, 00:06:23.592 --> 00:06:27.178 align:center line:-1 position:50% size:55% which the non-monotonic edge is slower. 00:06:27.178 --> 00:06:31.266 align:center line:-1 position:50% size:36% Right now we're set up on Channel 1, rising edge. 00:06:31.266 --> 00:06:36.396 align:center line:-1 position:50% size:55% If it occurs for greater than, about 120 ns, 00:06:36.396 --> 00:06:43.987 align:center line:-1 position:50% size:42% so let's increase the time until we get up to about 130 ns, 00:06:43.987 --> 00:06:50.785 align:center line:-1 position:50% size:50% and there you can see that it locks on to the non-monotonic edge. 00:06:50.785 --> 00:06:56.124 align:center line:-1 position:50% size:45% Now again at this point, we've isolated the problem signal, 00:06:56.124 --> 00:06:58.627 align:center line:-1 position:50% size:54% maybe use other channels of the scope, 00:06:58.627 --> 00:07:03.465 align:center line:-1 position:50% size:42% maybe at the time of this edge, something else is occurring. 00:07:03.465 --> 00:07:08.637 align:center line:-1 position:50% size:54% Find what that other thing is that might be coupling in and causing this problem, 00:07:08.637 --> 00:07:15.560 align:center line:-1 position:50% size:61% and it was the advanced parametric triggering that allowed us to isolate this edge. 00:07:15.560 --> 00:07:20.106 align:center line:-1 position:50% size:51% You've just seen a few examples of using the scope's advanced triggering 00:07:20.106 --> 00:07:25.862 align:center line:-1 position:50% size:66% to synchronize acquisitions and waveform display on some fairly complex signals. 00:07:25.862 --> 00:07:31.451 align:center line:-1 position:50% size:53% In this case, a random glitch and an infrequent non-monotonic edge. 00:07:31.451 --> 00:07:34.954 align:center line:-1 position:50% size:52% I quickly showed you some of the other advanced trigger selections, 00:07:34.954 --> 00:07:37.707 align:center line:-1 position:50% size:60% but we didn't have time to go into any details. 00:07:37.707 --> 00:07:40.585 align:center line:-1 position:50% size:46% One of those was serial triggering. 00:07:40.585 --> 00:07:45.340 align:center line:-1 position:50% size:54% In our next lesson, we'll be talking about serial bus analysis, 00:07:45.340 --> 00:07:46.883 align:center line:-1 position:50% size:46% which may come in handy for you, 00:07:46.883 --> 00:07:50.428 align:center line:-1 position:50% size:43% especially when you get to your capstone design project. 00:07:50.428 --> 00:07:55.100 align:center line:-1 position:50% size:63% If you're interested in learning more about the importance of waveform update rate, 00:07:55.100 --> 00:07:58.311 align:center line:-1 position:50% size:43% which we also briefly mentioned during this lesson, 00:07:58.311 --> 00:08:00.522 align:center line:-1 position:50% size:47% go to the URL listed on your screen 00:08:00.522 --> 00:08:03.858 align:center line:-1 position:50% size:60% to download a paper I authored on this topic. 00:08:03.858 --> 00:08:08.988 align:center line:-1 position:50% size:48% It compares the odds of capturing something infrequent to rolling dice. 00:08:08.988 --> 00:08:10.949 align:center line:-1 position:50% size:30% See you in Lesson 14. 00:08:10.949 --> 00:08:13.743 align:center line:-1 position:50% size:25% Go, Stanford Tree! 00:08:13.743 --> 00:08:15.286 align:center line:-1 position:50% size:19% A tree, really?