12

So i've been watching the MIT 6.002x Classes and they are really interesting, I feel like i've got a good grasp of Circuits and the Basics (I graduated with a CS degree...but EE appeals to me too).

Anyways I noticed....probably like a lot of people that a lot of the advanced math isn't really used on the job. It happens...better to know it than not I guess. But leaving out Signal Processing and similar "Intense"/math heavy sub-fields of electronics....how much Advanced Math do you feel like you use on the job?

I guess someone that did circuit design, Microcontroller Programming and such how much math would they really get into.

And Second Question: Is there a book that goes into the Advanced Math that's necessary for this? or do "most" electronics books pretty much already have what is necessary.

  • 2
    The only maths I use when it comes to Current Calculations and Heat dissipation... – Swanand Oct 09 '12 at 05:14
  • circuit design is an art more than a methodology. There are math in VLSI design and simulation(s). These days we use electromagnetic simulation tools.There I heard lots of math there. But when it comes to electrical, there are definitely tons of math + electromagnetic theories. Simply you can't explain how a machine works without PDE. – Standard Sandun Oct 09 '12 at 05:26
  • 2
    When working in electronics design I did not have to use any advanced maths in most of my day to day activities. But on occasion something would come up that required it. So it is good to have the higher maths knowledge, its another tool that a lot of people don't have. There is an infinite world of math though so I would start with probability and PDEs. Plus even if you are just doing circuit and MCU programming, alot of the times that purpose of what you are building is to measure different phenomena, and that will require some maths to design an appropriate sensing system. – geometrikal Oct 09 '12 at 07:12
  • 1
    what do you mean by "advanced math"? A bit of calculus that's in MIT 6.002x Classes is not that advanced, atleast for me. BTW, i'm a CS too ;) – miceuz Oct 09 '12 at 07:44
  • 3
    I've found that electronics design is 40% searching parts catalogs and datasheets, and 40% double-checking diagrams, with only a little bit of actual maths in the middle. – pjc50 Oct 09 '12 at 12:54
  • Seems that way....I mean obviously we use some basic algebra and such for figuring out voltages and whathaveyou....but besides that I don't see much Advanced Calculus or anything pop up on here. –  Oct 09 '12 at 14:52
  • 1
    More knowledge is always better. If you have time and motivation, why not? When we have more tools, we can build more and better things! – m3dl Oct 09 '12 at 17:03
  • 1
    Sorry, I am sure all will be super happy to see this, but this is the type of list question that is considered a bad fit for stackexchange. This type of list/fun question is not what SE is about. – Kortuk Oct 09 '12 at 20:33

5 Answers5

21

The ordinary common stuff is mostly just basic algebra, like Ohm's law, computing one of frequency, resistance, and capacitance from the other two, etc. The important skill here is not so much math but intuitively understanding the physics behind what you are doing. If you can look at a schematic and feel the voltages pushing and currents flowing and how each of the parts react to those, you can pretty much derive the equations you need to quantify things.

I also find basic physics is very useful to EEs, at least the kind of EE I do, which is designing small embedded systems. My job never just ends at the circuit or the firmware. To do the job right, which is to solve the problem not just make the circuit work, you have to have a good grasp of whatever the circuit is controlling or measuring. This requires good understanding of the system and the physics behind it.

All too often you find the people that know the system and wrote the requirements for what your controller is supposed to do don't really have a good grasp of what things are reasonably possible. They think up one means to solve the problem, then specify a circuit to do that. In other words, they know their world, but don't know yours very well. It is very valuable if you can be the one to reach accross (because they can't or won't), look at the big picture, and propose a better method of solving the overall problem. However, you can only do this if you have a good understanding of the system, which usually requires good basic physics skills on your part.

This brings up another important skill of being a good engineer, which is surprisingly rare. Always take the time to understand the bigger system your little design fits into, then look at the big picture. I find people are usually more than happy to talk about how their part of the system works, so go around and learn it. Then look at the overall picture and see if what you were asked to do still makes sense or only from the point of view of the one guy your gizmo interfaces with and that guy was only looking at his isolated problem. You may think this is a no-brainer, but then you'd be surprised how often this happens, particularly at large companies. The type of people that like to take a narrow view and work on just their little problem tend to gravitate towards large companies. There is room for people like that in a large project, having a few of them in the right place is actually useful, but it takes a skilled chief engineer to utilize these and all the people properly. That last part is very rare nowadays, and you will often find Joe Blinders in charge of stuff he shouldn't be. Even if Joe does try to look around a bit, he often doesn't know what electronics can and can't easily do. The worst is when he fancies himself a EE but doesn't really know what he's doing.

As far as more advanced math than common algebra, definitely learn to think in frequency space. I have done detailed frequency to/from time domain calculations on a few times, but the concept are valuable often. Every EE needs to be able to visualize what the frequency implications are of a time domain signal and vice versa. Here I am not talking about sitting down and solving Fourier transforms, but having a good intuitive sense of it. For me that came from doing the detailed math in college. I have done that math only rarely since, but the understanding behind it is useful every day.

Olin Lathrop
  • 310,974
  • 36
  • 428
  • 915
  • Would you mind going into some Simple Detail on some types of Physics you use? (you don't have to be specific..just some general stuff) –  Oct 10 '12 at 04:14
  • 1
    @Sauron: I am a consultant, so work on a wide variety of products accross many different industries and markets. There have been many example. Understanding something about electron beam tubes helped a lot in a kicking around ideas for a xray tube controller. Understanding something about fluied flow helped in the pressure controller circuit. Human vision insight was useful for the LED sign design. Flight physics was very useful for the airplane simulator. All these were actually out of my specific problem area, but the insights to other areas were very valuable. – Olin Lathrop Oct 10 '12 at 13:28
11

I find I use mostly simple Algebra day to day. Calculating power consumption, currents, resistor values, and thermal issues. For everyday practical circuit design like you're talking about it's more about creative problem solving than math. I'd take a guy who was a good debugger over a good mathematician any day ;)

That being said there are days when it comes in handy, you might be asked to design a system that requires higher level math to understand. It's usually around some control, communication, or signal processing problem (for me anyway). I can think of one example where I was designing a PWM audio output but it sounded "crackily". It wasn't until I read some papers, and used some matlab to do some sum of sincs that I was able to clean up the sound.

Certainly there is a lot of advanced math behind the tools we use, such as EM field solvers for things like Signal Integrity Analysis, spice, and other modeling.

I have friends working on ASICs that take algorithms from the "math guys" and put them into ASIC form, there's a fair bit of math involved there.

You'll probably find more physics type math in the advanced robotics sector but again that's more about control systems.

I'm sure there are many more places I haven't thought of but in general I find that every day there's not so much math. When there is I can usually turn to one of many reference books to find the equation I need.

Some Hardware Guy
  • 15,815
  • 1
  • 31
  • 44
6

I do circuit design, microcontroller programming, and 1-1000 kW power electronics design. I've done some pretty complex algebra to derive converter system gain equations at times. Basic algebra to implement calibration routines for A/D values. Calculus was needed to compute the average current through a phase-controlled rectifier while charging a capacitor. Constant-power discharge of a non-ideal capacitor was a big ugly non-linear differential equation. Trying to analyze ringing in a switch-mode supply was four big ugly ones. (Still working on that.) And estimating the losses in a high-frequency switch-mode converter involved a couple simple integrals.

That's probably most of what I've done in five years, and I gather I'm doing more calculus than most. 98% of what I do does not require complex math. The other 2%, I'm probably the best-equipped in the company to handle, so it's definitely a worthwhile skill. What's most important is probably not the obscure details of how to solve every possible kind of equation. You can look that kind of stuff up. What's more important is understanding the fundamental concepts of it all. Just what is an integral? How do I make use of one? How, generally, is one set up? And what resources do I have or need to evaluate it once it's set up?

Also, having that understanding makes you confident that you can compute things, and that the universe really does make sense. I personally find that kind of confidence very helpful, sometimes moreso than the actual results of the equations.

Stephen Collings
  • 17,373
  • 17
  • 92
  • 180
4

I'm not sure what is meant by advanced math in context. But on a daily basis I use PDE's, calculus (including line integrals) and when preparing papers for publication there can be some very heavy lifting, and sometimes using math to develop new analysis/models of systems. But on a day to day basis I'll use Mechanical engineering (beam bending), Heat flow, Semiconductor modelling, Quantum mechanics, optics, transistor theory, circuit theory etc. so a real grab bag of different fields which are surprisingly similar. I tend to be more towards research side of things now and am brought in for solving critical problems in front line production problems.

placeholder
  • 29,982
  • 10
  • 63
  • 110
2

Most of the advanced math was taken care of by the scientists and engineers who developed the parts we put together so the advanced math is not necessary on our part in many cases. We strictly do the engineering side of things where advanced math isn't always needed cause they took care of that already and supplied us with the needed data to interface all the parts.

If one wanted to stay involved in advanced math, you are more likely to use it in transistor and IC design than soldering those parts together to make a circuit.

Rob
  • 344
  • 1
  • 2
  • 14