electrical engineering requirements

[mutz]

hi,
was wondering what are the requirements for anyone to be able to study EE university course, to be able to fully understand the operation of microcontrollers, processors, and they're varieties, and be able to design as well.
whether people who get to study it are all ready familiar with basic computing,
whether anyone who attend these courses has all ready have some background with PC's, understanding of memory operation, CPU operation, the software (not from a programming aspect), some basic/advanced knowledge of the CPU market, whether MB design is also included, etc.
cause without it, it imply seems impossible studying all of it in 3 years.

it takes a lot of time to get the hold of it, understanding what exactly is a computer, it's capabilities, the way it function, the logic behind it, the research, applications etc.,
one generally feels overwhelmed with the knowledge and amount of research which has been and is being done and which he/she faces when they begin.

 

[Cogman]

Please note that I'm a senior Computer engineering bachelor program and not EE (Computer engineering is EEs shunned brother We focus more on the digital design while EE spend quite a bit more time dealing with analog stuff). Either way, I'll try to answer your questions the best I can.

1. What are the requirements to be able to study EE university course?
Not much, I've seen people that know NOTHING about how a computer operates do pretty well in my CompE classes. While having some background in circuit design definitely helps, most courses are built to let you learn it in university rather then learn it anew.

2. What are the requirements to be able to design and fully understand the operation of microcontrollers, processors, and they're varieties.
Before classes? Nothing. The classes that are needed, however, is a different matter. It really depends on WHAT your going into. To operate a microcontroller takes little more knowledge of knowing programming and how to read the documentation for said microcontroller. To build the microcontroller, requires knowledge in digital design (Some HDL) and preferably knowledge of transistor design and implementation, and all the math that goes with that...

3. Are people who get to study it are all ready familiar with basic computing.
Most are as it is what usually interests them in the field in the first place. Some aren't. (I had one guy that didn't realize that visual studios saved files in actual folders on his computer, and didn't just pull them from thin air. Very frustrating asking him "Where are your files saved at?" and getting "I don't know, its just in the editor!").

Either way, if you are going to be good at it, you'll probably pick up basic computer skills fairly quickly. Most schools do offer classes for it. However, I haven't seen the ability/inability of operating a computer as being a big plus/minus for this degree.

4. Whether anyone who attend these courses has all ready have some background with PC's, understanding of memory operation, CPU operation, the software (not from a programming aspect), some basic/advanced knowledge of the CPU market, whether MB design is also included, etc?

Some of this stuff is quite complex and spans multiple classes (CPU operation), others is fairly basic, and covered in like a week (memory operation). Some isn't related (MB design, that has to do with PCB layout, which CompEs generally don't care about A lacky does that.)

We don't really talk about the CPU market in my classes, They just aren't all that important when talking about microprocessor design (Other then to mention that Intel holds the patent on such and such).


If I was going to pick one word that describes what a Computer Engineer does, it is abstraction. We first learn how things work at a descrete component level and then build up until we are working with complex systems. By breaking everything up into their individual components, we can easily conceptualize what is going on. (It does take a while describing each and how it works, but it is doable.)

For a Bachelors, 3 years simply isn't enough, Most schools are actual calling it a 5 year degree.

 

[mutz]

thanks cogman,
that's the idea.

 

[CountZero]

hi,
was wondering what are the requirements for anyone to be able to study EE university course, to be able to fully understand the operation of microcontrollers, processors, and they're varieties, and be able to design as well.
whether people who get to study it are all ready familiar with basic computing,
whether anyone who attend these courses has all ready have some background with PC's, understanding of memory operation, CPU operation, the software (not from a programming aspect), some basic/advanced knowledge of the CPU market, whether MB design is also included, etc.
cause without it, it imply seems impossible studying all of it in 3 years.

it takes a lot of time to get the hold of it, understanding what exactly is a computer, it's capabilities, the way it function, the logic behind it, the research, applications etc.,
one generally feels overwhelmed with the knowledge and amount of research which has been and is being done and which he/she faces when they begin.

A few things to remember. You are going to school to learn this stuff the expectation should be that you are there to learn this stuff. Remember to take it step by step. Every step will seem complicated but as you build up looking back you won't quite remember why it seemed so hard. For example my very earliest EE classes with just resistors and a couple of ideas of different models seemed hard at the time but it is second nature by now. Whether your school approaches it tops down or bottoms up it will be a step at a time and every previous class should be setup to prepare you for the next. At senior/grad school level you will be able to likely read research papers and grasp 95% of whats important.

Where I work we design high end server chipsets and the vast majority of our hires are MS not BS. It is very difficult to get enough depth *and* breadth in a BS (four year degree).

 

[pitz]

Yeah, basically, the others have covered it. You go to school for 4-7 years, and then maybe, just *maybe*, someone will hire you. Unemployment rate for EE's is very high, and its pretty typical for a typical job to receive 200-300 resumes, sometimes more. And heaven forbid, if you graduate and don't manage to find a job within a year (very common), then firms pretty much toss your resume straight into the garbage and you may as well just go kill yourself because all your expensive education and student loans have gone to waste (engineers have a notoriously difficult time getting jobs outside of the field).

Its not so much that people at the BS level don't understand things as well as at the MS level -- rather, the MS degree is more of a screening device more than anything.

I think you'll find in CompE that very little is actually spoonfed to you. You'll just be expected to 'know' the C programming language pretty much overnight; its not like there's a semester of coursework to get you up to speed on that. Its all very fast paced, and very easy to fall behind in.

IMHO, the real action is in software these days, not hardware.

 

[CountZero]

I think you'll find in CompE that very little is actually spoonfed to you. You'll just be expected to 'know' the C programming language pretty much overnight; its not like there's a semester of coursework to get you up to speed on that. Its all very fast paced, and very easy to fall behind in.

Counter point, at my school 2 quarters of programming that start at the absolute basics is required to even apply to get into EE, CS or CE. IMHO only a completely crappy program would expect any knowledge that isn't part of the curriculum or fairly easy to pick up.

 

[pitz]

Counter point, at my school 2 quarters of programming that start at the absolute basics is required to even apply to get into EE, CS or CE. IMHO only a completely crappy program would expect any knowledge that isn't part of the curriculum or fairly easy to pick up.

Yeah, that's probably a good idea, although, I learned C over the summer, and they offered a few optional tutorials in the first couple weeks of the semester, so it was no big deal. Don't know if I would characterize such an approach as 'crappy' -- there's only a limited amount of time for everything!

 

[mutz]

there's time for everything pitz, and when you do something, you better do it because you are interested at studying it, and not because you would like to work in it,
that would save you the suicide part ,
another point, is that you would be much more content in something your interested in, and so you will simply give it the best of your self, you would be good at it, and depending on how serious you are, you would be the best there is,
not from a competitive pov, as if you love what you are doing, you will flower in it.

i'm not planning on going to any school, i'm more of planning to study it alone, i think it is possible, and depending on one's determination and will to study, i think he can do it in a shorter then a university course period and even profounder (not trying to be arrogant or anything).

Every step will seem complicated but as you build up looking back you won't quite remember why it seemed so hard.

this is something that sparked recently,
almost a year ago, when first visiting AT HT forum, everything here seemed way up in the sky, people were talking in a language that was extremely difficult to understand,
today, it (well as much seen in the last days), it is became a normal talk, between normal people,
and back then, it was complicated as hell.
not to mention Wiki pages, on computer stuff, which were once seemed like a totally taken from another world, PDF's and anands reviews lol, today, they are readable,
scientific work, etc.
of course to a certain extent.
at first you struggle with every word, you get inside a link and follows another and another and so on and on, till your mind blows, and you forget where you started...

the main issue with computers, if you are new to it, is the amount of data that is being poured at you at the beginning,
some of you might have forgotten that already, or either others have been dealing so long with it, that they just learned it all with time.

rubycon has said in one thread which was opened here not long ago, that computers were much more complicated few years ago, and actually doubted it,
there is so much that there is to learn, it's like a pyramid when at the beginning you have to learn a terribly vast amount of data, and slowly, you start connecting between the ends and start to build a picture of the system and the entire computing "sphere".
havn't had the chance to deal with computers few years back, yet it seems, as much as it goes on and on, things become much more complex, structured, complicated, yet one start to figure things that he/she couldn't before, simply because he has more varieties of understandings.

 

[schenley101]

I don't know if you can learn it all by yourself. I'm an EE and i thought i knew a whole lot about computers and electronics before i went to school, but that was just the tip of the iceberg. lots of programing and math.

 

[Cogman]

I don't know if you can learn it all by yourself. I'm an EE and i thought i knew a whole lot about computers and electronics before i went to school, but that was just the tip of the iceberg. lots of programing and math.

the problem is, You don't know what you don't know until you take classes. I never would have learned what an op-amp is on my own.

 

[mutz]

yeah,
it ain't easy,
been dealing a lot with hardware the last 2 years, and every time you feel you really understand something, you meet someone, a research or a forum, where you simply feel like a total noob.
it happens a lot, through many stages, and each takes you somewhere.

not sure if learning to design chips the way AMD/Intel does is so attainable, even to CE graduates (don't know really),
there must be some background in the field, some "qualification" years,
at least, one must deal with it intensely, as in programing, to figure out idea's he/she had in mind which they discover to be untrue when actually testing them,
it all looks easy at the beginning, yet when getting further into it, it gets so complex some times, that you really don't know if you are going to make it.

the only thing that is a pain with university, is the fact, you have certain time to learn the degree, you have to study some things very fast, and others slowly,
(yet it seems to be dependent on the program too by what you were saying).
you can't approach the teacher with every question, then you have to wait to the end of the class, if you miss something, then you have to go over it while the course still goes on, and then you miss another and you get into a circle.

yet, you have access to labs, to equipment, to lectures, data, you take tours probably,
you don't have to figure out everything by your self, so it has some benefits,
other things, is info you don't always get through the internet..
if you join the IEEE, you have to pay for almost every document, sometimes it is terribly hard finding the right info..

when you are in a class, you don't really have to worry about it.

 

[Modelworks]

Be sure that what you are going to study is what you want to do. What I mean by that is often people take EE thinking they are going to be doing lots of hands on stuff, working with tools, soldering, scopes, etc and while that is true you often spend more time working on the computer and doing math. I wouldn't recommend the EE field to someone who enjoys building a pc and working with hardware unless they really loved math. It is kind of like someone who enjoys building homes as a carpenter trying to become an architect. They are similar but one is much more hands on.

I advise you to pick up the book The Art of Electronics , pretty much an industry standard and see if you like the stuff it is discussing and if the math is at a level you are comfortable with learning. If you like what you see go ahead and start learning that book and you will be WAY ahead of others in the class.

Here is an online version:
http://books.google.com/books?id=bk...&resnum=5&ved=0CCcQ6AEwBA#v=onepage&q&f=false

 

[CountZero]

Yeah, that's probably a good idea, although, I learned C over the summer, and they offered a few optional tutorials in the first couple weeks of the semester, so it was no big deal. Don't know if I would characterize such an approach as 'crappy' -- there's only a limited amount of time for everything!

Crappy was a bit harsh The classes we had were more about learning programming concepts but obviously you learn language constructs as part of that. As an EE I thought it an odd requirement for those that were going into non-digital stuff but in retrospect it seems like a great idea. You know everyone has this baseline so at worst you might have to worry about students learning a different language but the concepts are already there.

You can learn to design chips the way AMD and Intel do. The more surprising thing to learn is that the way they do it might not be the best way to do it, it just happens to be the way that works for them. The other piece is that rarely does anyone know every part in depth. In my experience very experienced technical people are either ridiculously in depth on one or two things or have a broad but shallow understanding of everything and you need both sets of people. I have yet to meet anyone that has a very deep and broad knowledge base while I'm sure the exist I think they are the exception.

 

[mutz]

You can learn to design chips the way AMD and Intel do. The more surprising thing to learn is that the way they do it might not be the best way to do it, it just happens to be the way that works for them. The other piece is that rarely does anyone know every part in depth. In my experience very experienced technical people are either ridiculously in depth on one or two things or have a broad but shallow understanding of everything and you need both sets of people. I have yet to meet anyone that has a very deep and broad knowledge base while I'm sure the exist I think they are the exception.

really know what you are talking about,
i think, in order to understand computers (or what you are doing in general) good, you have to know everything there is to know about it, and i don't think they pass you this feeling or knowledge at the universities, as there are many constrains.
was talking with an expert about some broadband connection implemented in japan a year ago, querying about the feasibility of such project with the current infrastructure, and the guy said, "i'm expert with CPU's, i know nothing about what is happening there and how it should work",
well, told him, if you call your self an expert, you better know a lot about every
computer related stuff, or either, simply don't call yourself so..

i'm might be talking here a bit high, yet i think, if one wants to work in any field, he really has to dig it to the core, and not satisfy himself with this area or the other,
this is exactly it, it's either most people study a specific and know crap about the adjacent one, or either they know a lot of general stuff, and know nothing in dept.
it's hard finding people who are really interested in what they are doing and not doing it as a comfort,
these few that do, are truly interested in the field they are working in, and bring the best out of it.

modelworks,
thanks for the link, i'm more into Computer engineering and not EE as it seems, as the guys here has told/are telling it is the study of designing and/or understanding the implementation of microchips/controllers and microarchitecture.
i'm more into free study and less for going over full books,
picking up information from different resources etc.
though if it good and well written, i'll have it deeply looked.

 

[CountZero]

really know what you are talking about,
i think, in order to understand computers (or what you are doing in general) good, you have to know everything there is to know about it, and i don't think they pass you this feeling or knowledge at the universities, as there are many constrains.
was talking with an expert about some broadband connection implemented in japan a year ago, querying about the feasibility of such project with the current infrastructure, and the guy said, "i'm expert with CPU's, i know nothing about what is happening there and how it should work",
well, told him, if you call your self an expert, you better know a lot about every
computer related stuff, or either, simply don't call yourself so..

i'm might be talking here a bit high, yet i think, if one wants to work in any field, he really has to dig it to the core, and not satisfy himself with this area or the other,
this is exactly it, it's either most people study a specific and know crap about the adjacent one, or either they know a lot of general stuff, and know nothing in dept.
it's hard finding people who are really interested in what they are doing and not doing it as a comfort,
these few that do, are truly interested in the field they are working in, and bring the best out of it.

I disagree completely. In no way should it be expected to know everything, as a matter of fact it is probably impossible to know everything and actually meaningfully keep up with the latest technology. Your case is a perfect example, network infrastructure and CPU design are at best tangentially related and even things that are more related still shouldn't be required knowledge outside of a passing familiarity.

A high level CPU architect shouldn't, and doesn't, need to know device physics, high speed serdes operation, PLL/DLL, voltage regulation, power distribution, clock distribution, manufacture testing, etc etc. They may need to know of these things and will likely need to work with experts in these fields to determine the best course of action but they don't need to know all of them in depth.

If you've spent any time at all looking at this stuff it will come up fairly quickly that there are aspects you enjoy and aspects you don't. You will not want to spend 40+ hours a week on the aspects you don't and will thrive on the aspects you do. That is why you work in teams, there is sure to be someone that enjoys the part you don't. Put those people together and you get a stellar product make the one person who only enjoys part of it work on all of it and you'll likely get a mediocre product.

 

[mutz]

i see your point,
it brings this discussion to a deeper level which i don't think we should get into currently.

though, thanks for all your suggestions :thumbs:.

yet,
at a different level,
this is actually allready happening through different area's of science, where engineers from different sectors work together in order to achieve an unified task.
it doesn't mean you have to study everything in order to create an outstanding processor, and work solely on it, but rather as much as you get into things, you understand they're wayyy of function much better,
this happening all over the internet, people combine to understand different matter where alone, it would just be too much.
this is happening also at the distribute computing segment, where instead of building a multi million supercomputer which is capable of producing 2 Peta flops of calculation speed, you simply hook up people from all around the world where each one is doing a little bit, and in the end, this sums up to couple of time a machine that no-one almost would be able to hold/maintain by himself.

you see it when your dog gets ill, and you go over a medical forum and watches the symptoms in a thread,
well, you don't have to study medicine for it, yet you go over some wiki pages to get the general idea, and you can already confront the veterinarian with some questions and even ideas.

and that's what is beautiful about it, when we unite, we can do miraculous things together, yet when we are solitary, we are bound to suffering.

 

[PsiStar]

Math! All math up through del, dot, & cross product. By that time you will know all about partial diffi que ... and besides, it is only identities & understanding when to apply them anyway. Really. Not a big deal. Just do the damn home work *ASAP* & you are good.

At that time you will have experinece in the school's computer language du joir which may be a bit of C/C++ progressing to matlab (most likely). Matlab is very C-like but comes with numerous function libraries necessary for advanced level electrical engineering classes. With Matlab, the classes are not necessarily a piece of cake, but you won't get lost in the forest for the trees ... in other words you will not have time to worry over optimizing code or the specific algorithm. That could come later depending on your interest & desired specialty.

You will have had classes in mechanical engineering statics (watered down for non-MEs) as well as Chem E fluids (also watered down for non-Chem Es).

Think math. Be prepared to think nothing but math. Think in terms of what you will later call mathematical essays ... aka finals.

For EE there will also be class(es) in electromagnetic fields and waves. This is what you really want to get to. Schools that offer this ought to have campus wide access to software such as Ansoft (Ansys) and/or CST. This is the primo software that full commercial licenses cost ~$200K. Know one of these, just half a$$ .. and you will have great job offers and be highly sought after.

On the other hand if all you are interested in is wiring transistors a few ICs & triacs together ... expect lower pay.

 

[Howard]

Be smart and good at math.

 

[esun]

There's some really bad/misleading statements littered throughout this thread. I don't want to go point by point to correct all of them, but I just want to say take all given advice with a grain of salt (probably including my own).

Note that electrical engineering is a broad field. Extremely broad. People typically pick their specialties relatively early (I'd say by the 3rd year of college is typical) and only get a brief overview of other areas in EE. For example, someone that designs analog circuits will likely not be an expert in digital design, DSP, communication theory, CPU architecture, embedded systems, antenna/wave theory, microfabrication, or devices/solid state physics. All of those are technically EE, though (and I probably missed quite a few).

Most of those fields require the kind of understanding you would get from higher level maths, but not in the sense of having to know multivariable calculus or PDE particularly well (aside from antenna/wave theory and solid state physics). Higher level math also includes discrete math, DSP algorithms, and stochastic processes, which are very important in some areas of EE.

Oh, and ignore the scaremongering over the difficulty of EE and the supposed lack of jobs. For some reasons EEs love complaining about how hard it is, but if you're smart and work hard you'll get it. And the jobs situation...well, we've had a long discussion about that in another thread but suffice it to say there are plenty of jobs for EEs out there.

 

[Matthiasa]

Yeah just be good at math, and be able to study hard... Unlike me.
For math you will probably need calc 1-3, differential equations, most likely linear algebra, and the EE classes will also cover how to solve PDE(if not covered in your differential eqiations class) but you will get those in your physics class as well with a similar form that uses separable variables so little worry.

The analog stuff is also about 239048032 times harder then the digital/ programing stuff.
Really... even simple circuits end up with fun equations for them, but that applies to other engineering disciplines as well.

But yeah as mentioned just do your hw and study a lot and you will be set. Oh and you must be self motivated as well.

Oh and your gpa will likely be ~.4 or more lower then just about everyone you know just because thats how it works.

If you really want to know see if a local university allows you to ghost a student for a day or to get a feel of it.

 

[CP5670]

Math! All math up through del, dot, & cross product.

That's not asking for much at all. I see many people emphasizing the math that EE involves, but it should be put into perspective. It will also depend on what subfield of EE you are into. Some areas such as signal processing or control theory will involve more math than others, and in some fields there is also a big difference between undergraduate-level material and frontier research topics.

I'm not actually an EE myself but for the areas I just mentioned, multivariable calculus and linear algebra along with some elementary parts of various analysis fields (harmonic analysis, probability, complex variables and PDEs) should suffice. Many physics and engineering departments offer some sort of "mathematical methods" courses that cover this material quickly.

 

[mutz]

If you really want to know see if a local university allows you to ghost a student for a day or to get a feel of it.

that's a nice idea..

the thing to add here to your comments, is that one generally don't need to mess with all sort of equations unless he/she actually uses them,
like you never have understand using a debugger if you don't debug or do RCE.
it is all coming with time, and the matter is, as long as you don't actually need to use something, you simply won't find it relevant.
as an add, you would probably find it very hard studying something which you don't actually need, uninteresting, and probably even forgotten after a short while.

so everything should be taken step by step, when you have to figure out more complex stuff as a part of your work/project, you would simply gather the energy and interest in order to figure it out, while you will be working with it, you will understand or find the technique and bring it to a state of the art.

now only thing left, is to find the first clue, or slowly filter it from the scatter, catch it, and pull the string,
the rest will follow.

 

[PsiStar]

now only thing left, is to find the first clue, or slowly filter it from the scatter, catch it, and pull the string,
the rest will follow.

I would be a little more deliberate about it.

US News & World Report has a lot on schools

Most schools have course descriptions & the prerequisites listed online ... always worth a scan or 5.:\

Make sure that you go for what you really want. There is a difference between engineering science and engineering technology degrees. Engineering science degrees are mostly math majors with specialty, but still with serious labs.

ET degrees consider themselves more "practical" or "applied" engineering offering much lab time.

My take, if you are shaky with math then investigate the ET degrees closely. If you are an all thumbs klutz, then perhaps "hands on" might not be your best choice? Google is your friend on this.

After several years as an EE, I thoroughly enjoy the physics & the opportunity to invent. Tho, I rarely have the time to build my own circuits ... we do get to have a life.:awe:

 

[schenley101]

I would be a little more deliberate about it.

US News & World Report has a lot on schools

Most schools have course descriptions & the prerequisites listed online ... always worth a scan or 5.:\

Make sure that you go for what you really want. There is a difference between engineering science and engineering technology degrees. Engineering science degrees are mostly math majors with specialty, but still with serious labs.

ET degrees consider themselves more "practical" or "applied" engineering offering much lab time.

My take, if you are shaky with math then investigate the ET degrees closely. If you are an all thumbs klutz, then perhaps "hands on" might not be your best choice? Google is your friend on this.

After several years as an EE, I thoroughly enjoy the physics & the opportunity to invent. Tho, I rarely have the time to build my own circuits ... we do get to have a life.:awe:

Look at MIT's opencourseware site. they have all of the tests and lectures for their classes posted. for some odd reason, i like to try to do their exams for fun.

 

[schenley101]

and also, you end up learning all that math so you don't have to use it. thank goodness for Laplace and Fourier.