This article appeared in Evaluation Engineering and has been published here with permission.
The embedded electronics industry has been going through an extended period of disruptive evolutionary growth that is changing the way we do nearly everything. From advanced wide-bandgap semiconductors to machine learning and cloud computing, every powered device is becoming smarter, better, and more efficient.
Recently, Tektronix held its 2021 Tektronix Innovation Forum, which provided an opportunity to not only learn about the latest developments and future trends in high technology, it gave participants a chance to get a feel for how their peers are handling the latest technological developments. The online conference had presentations, panel discussions, and a keynote by former NASA astronaut and engineer, Daniel Tani.
We talked to Suchi Srinivasan, Product Line Manager at Tektronix, about the event and some of the trends and technology she saw, along with her insights about the current state of the electronics industry.
EE: Hi Suchi, one of the aspects of electronic design is that it's a double-edged sword. The more advanced the technologies are, the more advanced the test technology has to be in order to evaluate and measure it. Right?
Suchi Srinivasan: Glad to be here. With our customers, the tech evolution, whether it’s about ICs, modules, or systems, that evolution has requirements in terms of signal characteristics, what needs to be measured, that definitely increases the complexity of the test solution. The fact is that the test solution has to be much better when it comes to resolution, speed, and the ability to capture the right information and translate that into good data.
EE: The term I love to use is the Red Queen's race, which is from "Alice Through the Looking Glass." You're running as fast as you can just to stay in the same place. Because in test, the lines on your ruler always have to be closer together than the lines on their ruler. You could point your finger at any advanced developing technology, and it’s a test and evaluation challenge, too.
Srinivasan: Absolutely. Some of these market disruptors that you just talked about, like wide-bandgap semiconductors, for example. Now, here is a technology that really helps us improve our power efficiency, helps us scale so accurately with respect to size, but then with that comes the complexity of handling higher voltages, higher current, and timing characteristics, which are different from traditional transistors. So, innovation is definitely the name of the game for us at Tek, whether that means, we are innovating our power supplies, source management units, or signal management units.
In order to be able to handle that kind of power and current, when it comes to characterizing these wide-bandgap components, or scale that to thinking about system-level solutions, we are building up technologies for probing that can access these signals the right way, and translate the signals without losing that signal integrity. Our signal-acquisition systems within our scopes, it is important to have the right resolution in order to be able to capture those signals, and then the software being able to process these in the right way. Being able to capture the critical parametric and timing performance, and so on.
So we can scale from IC-level testing all the way to module testing and system-level testing, in order to help improve the time to market for our customers. There are similar examples of what's happening in other application areas like wireless, for example, 5G and 6G and IoT, and similarly, in the wired segment as well.
EE: No pun intended, there are a lot of moving parts in all of those application spaces, and they all have to be monitored. One of the aspects of the expansion of smart everything is, eventually everything that draws electrical power will be connected in some fashion. What are your thoughts on that?
Srinivasan: Absolutely right. So the connectivity space has definitely evolved. We're getting into times where our doorbell is able to communicate to our cell phones, our cell phones communicating to the lights in our homes. I've been able to talk to some of our smart speakers and essentially get things delivered at home. It's definitely true, we live in a world of convenience.
With that, the introduction of new technologies, whether that be, ultra-wideband, the new generation of Wi-Fi 6, and so on. And with this connected technology comes complexity of handling the higher bandwidth requirements. And that's where again, innovation within Tek plays a big role, because as we're continuing to stretch the capability of our processing units, we keep some of these standards in mind.
As we kind of scale to system level tests, for example, with ultra-wideband and some of those requirements when it comes to measuring signal integrity, and so on. Being able to scale our system-level application solutions to allow our customers to be able to test those solutions in time, and to be able to accelerate their time to market, is definitely top of mind for us at Tek.
EE: That makes all the sense in the world. Now, you recently had the Tek Innovation Forum, in fact, full disclosure, we participated in a panel during it. So, talking about the Innovation Forum, what are some of the things that you came away with from the forum? Do you have any interesting observations or insights?
Srinivasan: Yes, I was involved with some of the planning as well, and it was a great experience to kind of be a part of that, and see the evolution, with commentary from our customers and our field teams. I would say that a couple of standouts for me were around the comprehensive nature of the Tek Innovation Forum. The fact that we address a lot of these market-level disruptions, the evolution of technology within the different applications, whether that be power, wired, or wireless markets. And the next step of talking about innovation and technology within Tek, whether that be with our scopes, our probes, an accessory, or our software solutions, which kind of bring it all together and help our customers.
I think it was a nice, comprehensive event where we addressed some of the key issues that were involved. Then finally also thinking about the journey and evolution of our customer personas, the remote work from home situation as the world evolves.
Some of the tools that we are building out are in order to allow for connectivity, whether that'd be our cloud-based solutions, or even analysis with our TekScope, that kind of goes across and analyzes multiple channels, serving some of the needs that they're seeing in the industry. So I feel like we had this holistic approach of covering the markets, the solutions, the hardware, and the software solutions that we're providing, plus the tools that are available to our customers now, in order to get their overall solutions to the market on time.
EE: Getting to the market on time, cost effectiveness, and speed. Those are all very interesting aspects of test. But then again, those are things that have been around for a while within the industry as pressures, since automation was first begun, but there's an aspect of test that we've observed in the last, say decade or so. The term we could use, and feel free to use whatever vernacular you want to describe it, because there is no real term for it yet, but let’s call it product awareness.
For example, in the old-timey days, about 20 years ago, you could have an idea, and draw it on a piece of paper, or a bar napkin, or something. And then you’d go back to the office and look into a paper directory for various parts. Then you start by breadboarding some things, and then make a prototype. Then the team works on it for maybe six months to a year, then you send it off to somebody else to work on it for however long a period of time. Then that prototype is tested. And then it gets sent to a manufacturing facility and it's maybe tested at random for QC. And once it gets sent into the field, the only time you see it is if it comes in for servicing.
Nowadays, you can design in a simulated software environment, doing design on-the-fly, and your concept is being directly converted to a schematic. And the pandemic brought the perfect storm for collaborative software. So now we've got software simulation and collaborative design. Now it goes to prototyping. Today you can order modules from a distributor, or you can go to many manufacturers' websites and design your subsystems there, and you’ll get it all in the mail ready to assemble.
Then you’re manufacturing in a Six-Sigma facility, that's testing everything after every phase of manufacturing. And then depending on the product, you could be doing over-the-air software updates. So theoretically, you can be completely aware of a product from the moment you conceive of it, until the warranty runs out in the field. Today you can literally go to a product at any point in the chain, and know everything about it. A truck gets a flat tire with the parts on it, and you could see the numbers change on the spreadsheets. So now having said that huge setup, what are your thoughts on that concept of the ever-present awareness of product?
Srinivasan: A couple of things that you mentioned kind of resonated with me, because I've been in the technology industry for about 22 years. I actually came into the industry in engineering, and I've done board designs. I remember, as you said, drawing schematics, sheets of paper. And then trying to capture it, using a tool, and then going through that process of the layout, talking to the layout engineer, talking about the number of layers we need on a board, the signal integrity and so on. Then eventually kind of getting that into fabrication and the cycle timing was there, which is now, how quickly you could get something captured, simulated.
All those considerations. Is there enough signal spacing for signal integrity? What about noise, coupling, and so on? Now we quickly handle it when it comes to the simulation software that's available. Similarly, fabricating a board, getting a board back in-house in order to debug, and all the debug tools that are available. It's just night and day in terms of difference and the acceleration that we've had there. With all that, I think the big takeaway is, within the test and measurement industry, we have to stay on top of all this, and evolve quickly.
Typically, when I'm spending time with my team, we are doing these idea sessions in terms of what's next, and what are the big disruptors that are coming. The things that we always keep in mind are, how much do we understand the customer's pain point? What are they going through right now, that is preventing them from being able to debug faster and get validation of their solution?
I can think of a recent example, where we have these wide-bandgap semiconductors going in a big system, a generator board for one of our end customers. They were having problems with signal acquisition. Not only that, they were having overall system-level problems and they did not know what was causing it. So their mind directly went to, hey, it must be that wide-bandgap device, because that's the one thing that we changed on the board, and so on.
That's where, with Tek's help, given our probing solution, they were able to not only access those signals, but were also able to translate that with respect to the right requirements, voltage and current requirements. And our scopes were able to translate that into something that is visual. They were able to debug each part of their signal chain and figure it out where exactly the problem stood. So, that evolution in terms of understanding the customer needs, whether that be the debug methodology, or whether that be the software needs that they're looking for.
Even in terms of processing needs, we continue to see IC standards evolve. Multiple channels now are needed in terms of signal analysis, not just for the closing signals, but for the conditioning signals, the output signals and so on. So they had, again, another example with our evolution with the TekScope, where you can have four to five scopes, and synchronize all of these scopes, and look at multiple signals, all at the same time, in order to mix and match and analyze circuits.
Those are some examples of how Tek has been evolving quickly, given some of the needs that we see in the market and the acceleration that is naturally happening in the market space as well. Also, as part of these idea sessions, we think about some of the future problems. Data centers are a great example, we're all using tons of data, everything that we use in our daily life, mobile phones, computers, and now with autonomous driving as well, with cars starting to communicate to each other, there's a big need for the data. These data centers are going to be huge hubs of data transfer. And data throughput is going to become a big need. So they're again evolving in terms of AI or machine learning in order to upgrade their algorithms, their software applications.
EE: It's all variations on edge computing, and once you start talking about that level of power, machine learning is the logical step. But now to talk about old-timey engineering again, it's the same argument that was used in thin client, back when the server was in the basement and the terminal was on the desk upstairs. It's the same argument, decentralizing logic and decentralizing processing power. It just seems to go in waves as we have the enabling technologies to do so. But now we're in the final stages, as it were, of ultimate connectivity, no wires, nothing.
Srinivasan: Absolutely. No wires, lower power, smaller size, and all these characteristics, kind of coming together to form the solution. Which has been an interesting transformation because we almost have to rethink the way we end up acquiring signals or processing the signals. And that actually brings me to a topic about sensing and sensors.
We know that everything in the world around us is analog and the sensors are going to play such a major role, in terms of acquiring signals from the real world, and then helping us translate into the digital world, which is where most of the processing happens, the communication happens, and so on. The one thing I see is the evolution of the sensors, whether that be for medical purposes, medical sensors, sensing in patient monitoring or sensing in the automotive application. Especially with all the safety requirements and so on.
Or on the industrial side for safety and processing, one common need ends up being detecting these low-level signals that are coming in, small movements, small variations. And the capability to translate that into something meaningful that the system can then work with. So that brings up a really big challenge for us in the test and measurement space, because now with these signals becoming smaller and smaller than our equipment capability, whether that be accessories, our probing accessories or even our scopes that are capturing the signals and our benchtop equipment that are capturing the signals. Their capability needs to go in the direction of being able to not only send these signals, but be able to kind of accurately process them, removing the noise and being able to characterize and provide the right accurate information.
So, that is something we've been heavily focused on over the last couple of years at Tek, on improving that interface, as well as capturing the right signal integrity, using our industry-leading front end. Which essentially takes a ton of design knowledge and expertise in order to, not only design debug, but also kind of develop. So, that is something we have leveraged extensively in the current environment and we hope will be an advantage for Tek in the future, as signal conditioning and the sensor interface becomes a big need.
EE: Precision matters.
Srinivasan: Absolutely.
EE: Signal integrity, accuracy, speed, depth of capture, memory, it's all part and parcel of that. There comes a certain point in engineering and electronic design where it crosses that threshold from raw numbers into art.
Srinivasan: Right. The first challenge, of course, is as you said, with precision, is capturing information and then needing to process it. That's the other piece of the journey. We've come a long way there, I would say from just looking at things on the time domain to, okay, can we parallel process and can we get both the time domain and the frequency domain together. Which I think is a step in the right direction, but again it requires a lot of innovation, but it gives us the ability now to understand some of the signals in real-time, to be able to understand the performance and essentially the debug capabilities. Being able to do something like that, looking at time domain, looking at frequency domain, in parallel comes with design innovation.
We take a lot of these features for granted at times, but it all comes down to the engineering, the thought process that goes into it, the customer journey, as well as some of the needs when it comes to debugging these circuits and debugging with precision. At Tek, the other aspect that we are continuing to look at is, given some of the nature of the circuits and how they're evolving the system, what do we need to do in terms of processing the data, and providing the right information to the customers.
Then we go back a couple of steps further in terms of addressing some of the test needs in an automated fashion. So the customer then does not have to spend that time building their own algorithms, writing their own scripts in order to analyze the data and extract that right information. We live in a world where wireless technology is becoming ubiquitous. Given that, how can we scale with respect to the accuracy of our measurements, the processing of measurements, the features that we have on some of our platforms, whether that be hardware platforms, software platforms, that is all part of what we constantly think about, and we ideate around, and innovate around at Tek.