Episode 2: Electronic Circuits, VLSI and Undergraduate Education with Associate Professor Nathan Neihart

In this episode, our guest is Associate Professor Nathan Neihart from the Department of Electrical and Computer Engineering (ECpE) at Iowa State University (ISU). Here, we talk about his research in designing electronic circuits for Very Large Scale Integrated (VLSI) applications and Radio Frequency (RF) communication, along with student resources for undergraduate education within the ECpE department. This episode was conceptualized, recorded, edited, and produced by Santosh Pandey from the ECpE Department of Iowa State University. The transcript was prepared and edited by Yunsoo Park, Ankita and Richa Pathak from the ISU ECpE Department. The communications and digital hosting was handled by Kristin Clague from the ISU ECpE Department. The music was provided by madirfan from Pixabay (Track Title: Rainbow Machine).

Transcript

0:00

Welcome to our ECpE podcast series where we talk about exciting activities within our department. I'm your host, Santosh Pandey. Our guest today is Professor Nathan Neihart from the Department of Electrical and Computer Engineering at Iowa State University. Dr. Neihart, thank you for joining us in this episode. We want to talk about your research in VLSI, your leadership role within the department, the VLSI curriculum, the teaching labs and resources available for students in our department. To start with, could you describe some of your research projects in VLSI and some of your cherished accomplishments in the field? Sure. My research generally focuses in the area of high speed, analog mixed-signal circuit design, as well as sensor interface circuits. So analog signal processing to take signals from the sensing transducer into signals that can be usable by scientists to do whatever sensing that they're interested in. And right now, I've got kind of two or three main projects going. One of them is in the area of power amplifier design for power amplifiers that can process multiple carrier frequencies simultaneously. So right now, if you look inside of your cell phone, there's probably close to a dozen separate power amplifiers right now. So you've got the cell phone standards, which even there, there may be three or four bands, you have WiFi and Bluetooth and GPS and so on. And so what we're interested in doing is, instead of having 12 different power amplifiers in the phone, can we somehow have a single power amplifier that can handle multiple carriers simultaneously. And this will also be advantageous in 5G and Next-Gen 4G type communications. Because everyone wants higher and higher bandwidths. But the problem is that the spectrum that we're using is a finite resource. We really have a sweet spot for wireless communication below 10 gigahertz. And so we're not really going to be getting additional spectrum to have wider channels. So what the industry is doing is they have something called carrier aggregation, where instead of having one, you know, really wide channel, what you can do is you can have three or four smaller channels, but then combine them and have them behave kind of like one bigger channel. And that works well as long as you have multiple channels kind of right next to each other. But the next kind of step is to try to make use of channels that are widely spaced. And that becomes a lot more difficult from the RF front-end perspective, because generally speaking, RF engineers like to design, you know, single band, narrow band systems. There definitely are some wide band systems out there, but the bulk of wireless communication is done in small single band chunks scattered throughout the spectrum. So we're looking at that. And I'm also, like I said, looking at some sensor applications. So I have been working with people in VetMed and Chemical, Biological Engineering Department on different sensor interfaces. So there's a lot of people that build sensors. And when I say sensors, I'm talking about like the transducer, something that takes a chemical and creates some kind of a change in an electrical component, and that's fine. And they're very interesting and useful, but the bulk of research in these sensor designs and implementations utilize expensive benchtop equipment. So benchtop vector network analyzers, benchtop field analyzers, and these are all very expensive and not terribly portable and not really feasible for in-field applications. And so we have been working on designing circuits that can basically translate these very small changes in electrical parameters of the transducer into something that can be digitized and processed in a microcontroller or what not. And the last thing is I've had a couple of smaller projects with local companies that have been pretty interesting, and these really span the range of circuits for visible light communication. So trying to have ethernet type speeds using, let's say a diode, or I'm sorry, an LED to communicate data. And I've worked with companies on validating WiFi systems where they've had really high spurious responses. And they've wanted to talk about, you know, filter implementations or ways to reduce those. That's great. So besides some of the technical challenges that sensitivity for sensors, or let's say having a single power amplifier within a chipset, what are the other technical challenges in the field of VLSI that researchers are working on? So VLSI as a whole is a huge field. And so I guess I'll comment more on the challenges in the areas that I've been working. One of the big ones, to be honest with you is maybe 10 or 15 years ago, you could work in the field of, you know, let's say RF integrated circuit design, and you could work on Low Noise Amplifiers and make some pretty significant contributions, but because everything nowadays has some kind of wireless functionality to it. And I don't see that getting any less in the future, the field of wireless communication has just become so diverse that it takes a huge group of people, really with complimentary backgrounds to make a product, right? You need signal processing and you need communication theory and you need circuit design and you need device physics to some extent. So one of the challenges is just knowing a little bit about a large number of topics so that you can actually talk with these people and understand, not only what they're saying, but also be able to make yourself understandable to to use slightly different definitions with the same words. So there's that. But, you know, in my opinion, that's also one of the things that keeps it interesting, but technically, you know, the big thing is we're running out of bandwidth. So 5G is coming online, right? And they're moving to much higher frequencies for 5G. There's some standards that are written for 60 gigahertz. And when you start getting into those frequencies, you really start dealing with parasitics and transmission line effects for everything, right? You can't just have a discrete resistor and assume that it's going to be just a resistor now there's inductance and capacitance and everything. So being able to correctly model that and simulate that so that you have some idea of what is going to happen after fabrication is definitely a big, you know, area. And then being able to deliver power to the antennas at those frequencies is a big deal. And so the circuits and the technology that goes into that, there's a lot of work being done in that area as well. I think wireless circuit design is a pretty mature field and now it's really being application driven, but there's still a pretty significant amount of work just into getting power consumptions, where they need to be. And, you know, interestingly enough, linearity, which is something that not very many people really talk about if you look at say power amplifiers or receiver front-ends. Everyone talks about power consumption and efficiency, but linearity is becoming a much, much more important topic as well. So you're doing a lot of signal processing for things like pre-distortion and linearization and stuff like that as well. Right. So you talked about the importance of learning signal processing, VLSI, microelectronics altogether. So what are some of the courses offered within VLSI, both for undergraduate and graduate students? So our group, the VLSI group in our department, I guess in general, we actually have a lot of VLSI courses, especially if you compare us against other peer institutions. We have courses on filter design. We have courses on OpAmp design. We have courses on power management and they all offered every other year. So every two years, which is fine for a Ph.D. student, it's a little bit problematic sometimes for some Master students. But I would say that, compared to a lot of universities, we have a really healthy number of classes. And the other thing that's nice is can't think of a counter example. But I would wanna say that all of our VLSI classes have practical labs. So those might be in the area of circuit design using cadence. They might be in the area of actually building power management units using breadboards, or for example, in my microwave engineering class, we actually fabricate printed circuit boards for gigahertz amplifiers and oscillators, and then the students assemble them and test them at the end of the semester. So we have a large hands-on component. And I think to some extent, that's also the Achilles heel of VLSI because it just takes a lot of time but we have a large array of classes, which I think is pretty good. Right. On a follow up note, I was wondering within a semester, is it possible for a student or a class of students to go from design to simulation, to fabrication and testing - all within one semester? Is that possible? Well, so in my microwave engineering class, yes, but we're doing printed circuit board level designs. If you're talking about integrated circuit designs, probably not in one semester, but definitely in two semesters. So what we've done in the past is students will go through, for example, our EE 501, which is kind of our introduction to IC design and it's heavily focused on OpAmps. And so the students will do a design and they'll do a layout. And in the past they've been given the option to then submit their layout for fabrication. And one of the big problems with trying to do it in one semester is fabrication by itself takes, I don't know, 6 to 10 weeks, depending on the schedules. And you're just waiting, right? So when the chips come back, the students will typically register for one or two credits of independent study. And that's when they will either do the bare die testing where we can wafer probe the signals on the die directly, or it might come back packaged and the students will do kind of a printed circuit board design for putting together the test fixture. So definitely within two semesters, one semester probably not though. Yeah, definitely. I think having a package design would be easier to test on a breadboard rather than having dies and using a probe station. Yes, definitely. Yeah. What are the other equipment and facilities that are available in the VLSI teaching labs and the research labs, both in terms of hardware and simulation tools? So again, I would say that's another kind of high point for our department is the VLSI labs. And they, to some extent kind of double as the undergraduate circuits labs, but we end up replacing the hardware probably once every two or three years. So new oscilloscopes, function generators, power supplies, digital multimeters. We get brand new ones every several years. And, you know, I've seen some really nice equipment in our teaching labs. We also have some specialized hardware for testing RF circuits. So we have a network analyzer that goes up to 20 gigahertz, several RF signal generators. Those aren't generally open to students, but they're used in certain classes, as part of the teaching labs. And from a software perspective, we have basically all of the main circuit design packages. So we make a lot of use of Cadence, which is definitely the industry standard when it comes for IC design. And we also use Keysight, Analog Design Systems at ADS, which is sort of the industry standard for microwave designs. We've also used Microwave Office and HFSS for certain aspects of high speed circuit design, as well as some modeling. So we have a really robust design environment set up is very good at managing these types of tools. And we can back that up with very strong test facilities as well. That's great. Could you elaborate on the nature of your collaborations with industry partners, such as Skyworks and what skills are these companies looking for from our students? I've had several collaborations and they've really kind of run the gamut in terms of what we've been doing. So with Skyworks, you know, their primary business is power amplifier design. So we were validating some of our concurrent dual band designs using some of their processes. I've also worked with other companies that are local. Ag Leader is a local company. And probably the longest collaboration I've had though is with Skyworks. And it's actually turned in to be pretty fruitful, I think in both cases. So we've done a little bit of research with them, but the main collaboration that I have is kind of between them and my EE 514/414 class, which is my microwave engineering course. And that course has kind of become a required course for students interested in doing internships and co-ops with Skyworks. I've had people from Skyworks come, engineers come and give talks in my class. Every year, I have probably two or three students at a minimum that go directly from the EE 414/514 class to co-ops and internships. And the other day I was counting, and over the past 10 years, I think I've got something like between 25 and 30 students that have taken full-time positions with Skyworks. And it's really advantageous for Skyworks as well, because one of the problems that they were having before I joined Iowa State was that they would have new employees join their company. They would go through the first two years of training them, which is sort of the most expensive time for a company because the new employee doesn't really have enough knowledge and experience to really start being productive. And then those people would end up leaving Skyworks and basically going to the Coast, either the East Coast or the West Coast. And so they were losing a lot of people. And one of the big things they've gained, I think through collaborating with Iowa State is we have a large portion of students that have grown up in the Midwest. So they have family ties in the Midwest and are actually looking to stay in the Midwest. And I would say that the bulk of the students that have joined there from my class that I've known in the past, you know, 10 years are still there. So it seems to be working in the sense that they're finding, you know, a pool of students that are interested in staying in the Midwest, but still have the technical skills and abilities to be productive in these high tech fields. Right. That is great. So what is the value of pursuing internships in companies for our students? Yeah, interestingly, I didn't do an internship as a student myself. So, I'll first, just start by saying, it's definitely not required that you do an internship, but it does help on a couple of fronts. One, obviously it gives the company time to kind of see how you work, see how you fit in their team. It also gives the student some time to see, you know, how that company fits within them as well. Right? So like how many hours do they require? What is the company culture like? What are their, you know, potential future colleagues like? I think that's important, but it also, I think more than anything, it helps students see that what they're learning in class, which can sometimes really seem academic, how it gets applied to real world problems. And that's something that I think academics in general struggles with, because there's so much that we're trying to teach students in a relatively short period of time that we don't have the time to really sit down and say, look, this is how it all fits together. This is how this is all used by companies every day to solve problems or create products that people need and want. And seeing that in an internship, I think gives the student a lot of perspective into why it's important to learn the types of things that help them learn. For students in VLSI, how can they apply for these internships or co-ops or job openings? Obviously there's the career services, right? They have CyHire and many, many companies post positions there. I would also, you know, I think everyone knows about the career fair and our department also hosts information sessions, I think is what they're called, but near the time of the career fair, a number of companies come and they'll reserve a classroom and, you know, give a little information session. But the other thing that I would say that sometimes students may not be aware of is if they've taken a class that they've enjoyed and they've enjoyed topic, get to know the professor a little bit, because at least for me, I have a number of companies that reach out to me, looking for qualified students that are interested in very specifically, you know, microwave engineering. And those companies may not always be posting those positions publicly. They're generally reaching out to contacts that they know, seeing what the pool looks like, you know, and I can say even if they are posted on CyHire, an emailed CV from a faculty member that works with that company is gonna get a lot more attention than a CV just randomly submitted online. So talking to faculty members, I think, is a big step also. Within our department, we also have the student services that help students prepare their resume. Yeah, absolutely. Yep. Okay. How can our students reach out and participate in the VLSI? Yeah, that's another good question. And again, that comes through, I think personal relationships. Now everyone's different, but the times that I have hired undergrads have always been for me personally, undergraduates that have taken classes from me and have, you know, made an effort to come to office hours and, you know, come and talk to me about things that they're interested in. And that is generally how a lot of the undergraduate research opportunities that I've had personally have come up. It's a student that's talked about being interested in some area, and it turns out that that's also an area that I've been, you know, working in and have had some portion of that, that I can break off from a graduate student project. So I think, you know, going to office hours, talking to your faculty members can really get you ahead in a couple of different areas. Another question related to students. As you know, students have limited time to pursue multiple courses in multiple fields, but do you think VLSI and microelectronics are kind of integrated together and do you think, or do you suggest students should take both courses in VLSI and microelectronics or semiconductors? Yeah, it's a good question. I, you know, they're definitely interrelated, but, you know, I guess I would say that there's plenty of good VLSI designers that, you know, have a passing knowledge of semiconductors and there's plenty of good semiconductor engineers that have a passing, you know, knowledge of VLSI. The people that are the most successful though, in my opinion, are people that know a bit about all because, especially when we're doing high frequency design, it's hard to have a good design unless you really know how the system is fabricated because how it's fabricated dictates a lot of the parasitic capacitance and inductances that we have to worry about. And it's easy to think that you can just have this kind of black box and, you know, say, okay, yeah, there's some capacitance and stuff, but knowing a little bit about how it's physically put on the chip, I think can really provide a lot of design insights that would make you a better engineer. But like you said, we have limited time. So, you know, you kind of have to focus. You know, another thing I guess I might advise students is you've got a lot of time to specialize in your career. When you're at university, it's one of the few times where you can study something just for the sake of being interested in studying it. And, you know, maybe exploring a certain topic that you don't know a lot about. But you're better off having a broad knowledge base that you can build on when you decide what you want to do in industry, because it's hard to know what you don't know, if that makes sense. You know, you may have a preconception of some topic, but when you actually start taking a class in those topics, it may be something completely different. Right? Right. So on a related note, how can students have access to some of these circuit design, simulation tools virtually in a remote location, especially during the pandemic. So, if there are students enrolled in Iowa State, then we have a number of virtual machines and servers that they can remotely access that has all of the software that we would use in a class. If they're not students, then there's actually a pretty healthy ecosystem of open source circuit design software. So tools like LTspice that are free for circuit design and simulation. Tools like KiCad, which are really quite nice PCB design software that's also open source. It'll take a little bit of looking depending on what you're trying to do, but there's actually quite a lot out there. But for students, they just need to sign in using our VPN and they've got access to everything that on-campus students have access to. That's great. Moving on to our next topic. I understand you are the Director of Undergraduate Education in the department. Could you describe your responsibilities in this role and what are some of your accomplishments so far in this role? Sure. So as the Director of Undergraduate Education, the big thing that I do is I chair our department's Curriculum Committee. And there's a lot that goes into that. We handle student petitions. So, you know, for example, if a student needs to fulfill, let's say a computer engineering technical elective requirement, the department has a list of classes that would fulfill that. But every so often, there's a student that takes some other different class because it sort of fits better with their professional goals. And then they'll write a petition to say, you know, I took this class, it's not necessarily on that list, but can I count it for this tech elective or whatever? And so the curriculum committee makes decisions on those. Recently, we have also kind of taken over the prerequisite approval process because our department used to be pretty lax with prerequisites and they weren't really ever strongly enforced. And then, ABET started placing a little bit more importance on that. And so one of the things that we've done since I've been chair was to go through and review all of the prerequisites for all of our classes and contact faculty to basically say, look, are these are the prerequisites that you really want, or should they be something else? And now our committee also handles the approval of these prerequisite waivers. Some other things that we've done in the past five years since I've been chair is we've kind of revamped a little bit, the computer engineering curriculum. We've added some computational thinking type of requirements, which I think really helps things like algorithms and just being able to break problems down into steps that something like a computer can handle. We've added some different technical elective requirements for the computer engineering program. And more recently, we're looking at introducing some new professionalism and ethics requirements into the curriculum, and we're in the process of developing a new course for that, as well as kind of refining some other aspects of both the Computer Engineering, Electrical Engineering and Cyber Engineering curriculums. Okay. So I presume there would be different challenges for students in different major programs, both in terms of classroom instruction and engagement in teaching labs. Could you comment on how you manage the student needs in the different major programs? Yeah, that is a big process, honestly, and it really kind of comes through the composition of the Curriculum Committee. So the Curriculum Committee has 6 faculty members in addition to myself, and it's two members from Electrical Engineering, two members from Computer Engineering and two members now from Cyber Security Engineering, which is a newer program. It's these 6 faculty members. And then we have one student representative from really more than anything, it's the student members Because as faculty members, we have some idea of what we think is going on in these classes and these programs. And then the students can come in and say, well, actually, you know, that's not how it works. And so that has also been a pretty big resource as well. All right. So I understand within the Curriculum Committee, the curriculum within each major program should evolve over a few years to keep in pace with the 21st century challenges? So do you think our curriculum is evolving over time? You know, I do. Actually it's an interesting kind of you know, the field is definitely changing, right? Electrical engineering today is much different than electrical engineering was in say the 40s and 50s. One of the challenges really is that electrical engineering and computer engineering are getting bigger, if anything. Things that once were just science fiction are now becoming mainstream electrical engineering, like biomedical engineering and all of this stuff. So what the challenges is in finding what are the foundations, what are the fundamentals that students need? And the good thing is those aren't changing too fast. There are certain fundamentals like circuit design, signal processing, electromagnetics, calculus, algebra, that kind of stuff. That's not really changing all that much from, let's say the 50s until now. But where our department is evolving I think a lot is in the 400 level tech electives. Because as faculty come in and they're working in new areas or developing new research programs, a lot of times they will be proposing new courses and that's another role of the curriculum committee is to help faculty through that process. But that's where the students start to see the evolution of the field over time. And I think our department is pretty good at that actually. The challenge then is, you know, as you said earlier, we have all of these interesting tech electives that come up, but students still only have, you know, four or five years that they're trying to finish their degree in. And so you know, they're forced to kind of pick and choose which ones seem to be most related to them. Right. We also have a number of student advisors that are doing a great job advising our undergraduate students. In your role, do you work closely with the student advisors? I do. I work very closely with Student Services actually. Because honestly, they are the people that know all of the rules that students have to follow, you know. How many tech collective credits, what's the difference between a dual degree and a dual major? What types of credits can account for what? And honestly, without Student Services, I wouldn't be able to do my job as the Director of Undergraduate Education at all. Another area I work with them closely in is in awarding scholarships. So every semester, I sit down with Vicky Thorland-Oster, who's sort of the lead in Student Services, and we go through all of the applications and all of the different scholarships and it takes like three days, but we make all of the different awards. And so, you know, again, that's something that I wouldn't be able to do without them because they often times know most of our students personally. And, you know, they can give a lot of insight into things like awarding scholarships. Could you elaborate a little bit on what kind of scholarships are available for undergraduate students and how can students actually apply for these scholarships? What are the deadlines? So, unfortunately, I don't know the deadlines. I would refer them to Student Services. Okay. But I have to say that there are many, many more scholarships than I was aware of before I took this position. We probably award between electrical and computer engineering, maybe 30 to 40 scholarships every semester. Some of those, or, you know, one year, some of those are one semester. Some of those are three year type scholarship and they really run the gamut. The other thing that I realized since taking this position is when someone provides money for a scholarship, they get to set the stipulations on what type of student this gets awarded to. Right. And there are some very unusual scholarships out there in our department. The one that comes to mind, I forget the name of it, but it is meant to go towards students from Iowa cities of populations, like less than 3000, that are majoring in music and electrical engineering. The criteria is set by the sponsor themselves. Yes. Yep. Where are the details about all these scholarships? Are they on the websites? They are on the websites and you can access them actually through our Student Services website. And I also think that the university as a whole has a list of scholarships that are available to students. How can students reach out to you, the Student Services or the department in general to provide feedback on course instruction from time to time? You know, I think feedback is really important and that should be the role of the course evaluations at the end of every semester. You know, I think the best place for feedback is through the student's advisors. Students seem to find that the more comfortable route. I, you know, am always open to feedback if students want to send me an email or come to my office in Coover hall on the second floor. But I do hear a lot of feedback through advising. Another venue is every semester our department hosts a student open forum that I actually moderate where students can come and talk about whatever it is that they wanna talk about. The other thing that I would maybe recommend is one throughout my career is that a lot of times when someone is good at something, we assume that they know that they're good at something. And so we don't ever really mention it or bring it up. But if you think about it, the times that you've been, you know, Hey, you did a really good job teaching that class, or you're doing a great job hosting this podcast. You know, those types of things. It actually means quite a lot. So I would say that if a student takes a class where they're impressed with how the faculty member conducted the class - say something. Because it goes a long way because you know, a lot of times we think that people know that they're good at things, but I think in reality, we're all insecure about what we do and we're all afraid that we're not actually good at what we do. Yeah. You mean both positive and negative feedback? Definitely. I think positive feedback goes a long way as well. Right. You know, obviously we want to hear the things that we need to improve, but it goes a long way to hear things that are going well as well. Right. Moving forward, what is your vision for undergraduate education in our department and what are some of the exciting changes that we can look forward to? That's a good question. I see a lot of momentum building behind changing the structure of lectures and the structure of how we educate our students. One of the interesting things is that with a couple of exceptions, most of us don't actually have formal training in education. And there's a reason that people can get degrees in education because there's a lot of research and a lot of theory and techniques in that go into how to educate students and how to reach students and how to, you know, structure your classes. And I see that there's a lot of momentum building now behind trying to move away from traditional PowerPoint, 50 minute lectures, three days a week. One of the things I've done in my senior microwave engineering class is I've kind of moved to a flipped classroom style where most of the time in class is spent working through examples, working problems in small groups and in large groups. And I think it's pretty effective. I'm seeing this more and more now in different classes in our department. And I think that it's going to result in a more enjoyable experience for students and more effective learning for students as well. Right. Do you have any final of advice for students who want Design? The thing I would say is to really learn the fundamentals. Everything that we do as circuit designers really stems from what we learn in EE 201 and 230. And what makes someone a good circuit designer or not is really how well they know those fundamentals because the insight comes with experience and the intuition comes with experience, but it's all built on five or six main principles. I think that was a great discussion. Good. I hope this information will be useful for our students. Definitely. Yeah, me too. Thank you. Yep. Thanks. Thank you. Bye.