Akshay Mehta – Applications Engineer

11 Nov

Introduction

In today’s interview, I talk with Akshay Mehta, who works as an Applications Engineer at Texas Instruments. Applications Engineer performs a variety of roles based on the industry and company. They typically play an important role in ensuring the engineering success of a product and their role requires working with both engineering and product/business teams. Akshay works in an important area of engineering – power engineering. This is a field that finds applications in any industry you can think of be it consumer devices to huge oil refineries. Everyone needs power! Read on more about Akshay’s Engineering journey.

Akshay Mehta

Hemang: Akshay, can you tell me a brief about your background? Your engineering journey so far…

Akshay: Ok, so I did my undergrad studies at Ramrao Adik Institute of Technology in Bombay. I did my Bachelors in Instrumentation Engineering with kind of a focus on control systems and control valves, a very industrial approach to things. Frankly, I went to this field because I was getting a cheap entry into engineering without paying a high fee. I really didn’t have any idea about what instrumentation engineering is specifically. I lucked out in the sense that I was enjoying what I was studying and I could understand what was going on.

 

After finishing my Bachelors I was wondering what I should do and felt like I should go and get Masters because what I had studied, and jobs for those in the country were kind of limited. I wasn’t really getting any interview calls anyway. I waited for a while to get interview calls, it didn’t really work out. So I applied for MS (Masters of Science) programs at universities in the United States and got a call from University of Texas at Arlington, for Masters in Electrical Engineering program. The engineering program was very diverse, and it took me about a semester to really figure out what I really wanted to do. I took up various courses such as Random signal generation, different filters, different transforms, and so on. Then somewhere around on the end of the 1st semester I realized that this is not what I really wanted.

 

I took up analog courses like CMOS and a little bit RFIC’s and eventually I figured after all this, regular circuit design was more interesting to me. I took up 2 different courses, one was kind of advanced filtering and the other one was a digital approach to Kalman filtering. I did well in both of those courses and figured out that I wanted to do something with a focus on control systems; how they work and so on.

 

After engineering I was looking for jobs and I just happened to travel to California to visit my uncle. I was there just for a small while, when I got a call from National Semiconductors for an internship. So I actually got an internship after I finished my graduate program! About that time, it was actually easy to get a job but not quite too. May be because may be I had a kind of very wide, focus was not into a particular area, so may be that made it a bit difficult to get a job or get a call essentially. But I got a call for an interview call for an internship from National Semiconductors, and that was for a position that it required me to create simulation models in SPICE. I had learnt something like that in school in University of Texas, in HSPICE, I hadn’t really enjoyed that but when I talked to these guys, they said oh it’s much different from HSPICE, a different flavor of SPICE especially.

 

I tried it out and it was an internship so I really had nothing to lose. It was just to gain some experience even if it didn’t work out too. I actually liked it a lot and I finished my one-year internship. They gave me a full time position and I worked like that as an application. The task was to create analog simulation models for DC-DC converters or voltage integrators. Everybody uses DC-DC voltage regulators on every application; every little system needs them, which are portable or non-portable. So it was a pretty generic field, which required kind of the niche knowledge that I had. One of my colleagues said that a DC-DC regulator is basically a control system on steroids. So I had a good knowledge of control systems, I was able to use that and by then roughly the 1st year of internship, I had good knowledge of creating SPICE models, so, it worked out really well. After 3 years in that particular appointment, I moved on to silicon testing and silicon evaluation, which was something I always wanted to do. So it’s been pretty much like a dream.

 

 

Hemang: The last role, which you mentioned, the silicon testing and evaluation, is that still at National Semiconductors?

Akshay: In 2011, Texas Instruments bought National Semiconductors. So now, I work for Texas Instruments.

 

Hemang: So you’re still with the same group, it’s just acquired by Texas?

Akshay: Yeah so the whole Silicon Valley group is now called Silicon Valley Analog, a business unit of Texas Instruments. So whatever was called ex-National is now called Silicon Valley Analog.

 

Hemang: So very few analog engineers coming out these days, right?

Akshay: Oh, very few. Everybody wants to just go into digital. I don’t know why. I guess it’s kind of not so cool sounding.

 

 

Hemang: I mean in one of my roles I’ve worked on developing displays. So we had the display system module itself. To drive that, you need a bunch of analog circuits. We had one input supply coming in externally, you had to amplify it to the right level and regulate it to whole bunch of different DC levels. The guys who could do analog work were very few.

Akshay: That’s right. You need analog engineers for that and that knowledge is not very easily found these days.

 

Hemang: One thing that impressed me about, I worked with a couple of really smart analog engineers, is that, there was a few calculations which were required, you’d have to go and choose some capacitors or inductors, which they’ll recommend, but then in the process of product development, you would come across issues. The circuit would not work as expected although the calculations were correct. By instinct they’ll come up with some other value of the component.

Akshay: You know actually I feel that analog is something built on a basic that we had studied in school. Like way back when we used to study in 1st year college, 2nd year college, when we had learnt about capacitors, inductors, all that knowledge is being used, and it’s interesting to see that all that is what is required to become successful. That’s it. Everything else that we study in university is built up on that.

 

I have some of my colleagues who are veterans with 20-30 years of experience designing silicon, designing analog ICs. And just to listen to them talk is fabulous. It’s like how you saw with those guys; they’d pull up stuff out of their hair and say “yep, this is how it works!” J

 

 

Hemang: Do you still create simulation models?

Akshay: I do. My job now doesn’t really require me to do that, we have licenses for various kinds of SPICE software, there’s one called SIMPLIS, there’s one called PSPICE. I use Simplis and I like it a lot. Most of the work we do is voltage regulators, I used to work with LED displays too, power for LED’s. But nowadays I just do voltage regulators. The way it works is that marketing comes up with a requirement. As an example, they may say this market requires something like 12V regulator, or 3.35V reg, and these are the inputs, so how can we come up with a solution that is clearly efficient and stable over a variety of inputs and make it as easy as possible for the user to integrate it with their product?

 

From there we come up with possible circuits, possible designs and then we relay the information to our design engineers who basically work on the silicon. They come up with a silicon design and eventually the final product with just the IC. At that point we start working on validating the IC, or testing the IC. Every time you go to the bench, to test the IC, certain tests require a lot of time, and certain tests are, it’s easy to just simulate them sitting on your desk. So, I still like to do that. I create a model every time there’s a new product, I use that extensively, it helps me a lot, so I still work on SPICE too, along with what I do.

 

 

Hemang: So because you have some idea of the model, you know what to expect. A lot of the testing guys receive a spec from someone and approach the validation process as just as pass or fail. They may not have the skills or the desire to go into as to why it is not working.

Akshay: Yeah, and stability is another issue, like control systems like what I’s saying, you need to know exactly how to design control systems, for it to be stable over a variety of inputs and variety of temperatures and all that stuff. It’s easy to first make a first task in SPICE and then really going through that and testing it out.

 

Hemang: Let’s talk about analog engineers since there are a very numbers of them coming out of school. How is new talent encouraged?

Akshay: One method is through internships for current students. The interns spend some 2-3 months with the group and basically doing what the engineers ask them to do, so it’s a good learning for all them. When they go back to school, if they like what they’re doing, they’ll change their focus or continue their focus in whatever is required to get a full time position for this job.

 

 

Hemang: It’s fun to work with interns. They come with an open mind and are willing to do any engineering task. They are enthusiastic about every single thing.

Akshay: Exactly, mostly pretty much all of them that we’d experienced are really dedicated towards what they want to do, they go back with a very clear perspective of what needs to be done to get a job in this field. Unless they don’t like what they are doing; I mean that could happen too.

 

 

Hemang: Yeah it’s good if they realize that as well.

Akshay: Exactly, positive or negative feedback is pretty important.

 

 

Hemang: I met a couple of interns at some of the other groups, who’ve told me that I am doing this role, I kind of like it but I am not sure if I want to spend my full time hours at this. So because they’re interns they do get some amount of freedom to move around, here and there. I think groups also understand that.

Akshay: Yes, that’s true. There are some companies who take you on internship positions even if you have finished your school.
Hemang: It certainly has worked out well in your case.

Akshay: Yeah, it worked out well because I was able to test it out myself is it really for me or not. Mine was an exceptional condition; it was weird. I don’t have words to explain it. There wasn’t a guy who was in a similar situation I was in after me, so. I don’t know what that tells me, I was probably not that great or it was weird, whatever. J

 

Hemang: So if a current grad or undergrad student was interested to get into this field of work, what would you recommend say as courses they should take?

Akshay: For courses to get into Power engineering, you need all power related courses. So basic is analog circuit, I mean there are certain schools for example like Santa Clara University who have course which are very much dedicated to this kind of work and that’ll include power engineering course itself. I’ll say Control system design, that is also very important, and basic Filter design is also very important. I meant Analog filter design and not digital filter design. Although digital filter design is also not bad because it gives you skills in math, that is also very important.

 

Circuit design like analog circuit design is really the basis of everything. Currently, for example, I work in application engineering, so that’s a little zoomed out. I look at board level design right, product level designs. Then if a person is so dedicated or so much enjoying their job, they could get into IC design, which is one level down, and that requires a course like CMOS design and other IC design courses for that.

 

 

Hemang: What do you think about the future job prospects in this field?

Akshay: Oh this field can never run out of jobs, it will never run out of jobs. As long as there is electronics engineering there are jobs in power design. How cool is it? That’s another story. I mean what I’ve heard from people is that it is not one of the most cool things to do. I enjoy it, it gives me exposure to various other things to do, but there are not too many power supply engineers or power designers. So the competition is not that great. It’s quite likely that even if you do change jobs, are let go or whatever, you will find a job fairly soon.

 

 

Hemang: It’s one of those industries, which for sure it has to keep up every other innovation. Right?

Akshay: That’s true. All of your products require power. So you’re never going to run out of jobs. As a doctor would you ever worry about career safety? Not really.

 

 

Hemang: Akshay, thanks for this info. This was very, very helpful.

Akshay: My pleasure.

——————————

You can read more about Akshay at his LinkedIn page.

Print Friendly

No comments yet

Leave a Reply

CommentLuv badge