Thanks for participating in Fundamentals of Electrical Engineering. Hopefully, you appreciate the breadth of applications and ideas that are all called "electrical engineering." Many EEs never build circuits, working instead at what is called the systems level. That is the approach we have taken; hopefully you have gained a strong sense of modern electrical engineering. For example, the last week's results on channel capacity are especially important in communication engineering and in many other fields as well. I have personally worked on calculating the capacity of a group of neurons, which finds application in the design of prostheses and brain augmentation systems.

One more plea for the post-course survey. Even if you stopped watching videos a while ago, please do complete a survey. The survey for those who have "completed" the course applies if you have studied most or all of the videos.

It has been a great experience teaching FEE online for the first time. I'm going to do it again!

Sun 14 Apr 2013 10:44 PM CEST

The last week of Fundamentals of Electrical Engineering (FEE) provides a detailed comparison of an analog communication system with a digital one, both sending and receiving the same speech-like signal. Guess which one has better performance. We then turn to the crowning result of information theory: it is possible to send digital data through a noisy channel without error if the data rate is low enough. When Shannon revealed his result in 1948, it turned the communications world upside down and started the Information Era. You'll find our comparison of analog and digital communications scheme changes when that result is taken into account.

Available this week is the Post Course Survey, one intended for those actively following the course and another one those who are not. I really want to hear from both groups; what you think about FEE and how it could be improved. You will find a link to the Post Course Survey on the left-side navigation list on the course landing page. The survey is available for two weeks. While taking the survey, you will discover that I am trying to design a laboratory course that goes along with FEE. At Rice University, a lab is already integrated into the course and I want to try to do the same with the MOOC version. We have many options for the kind of laboratory experience we could provide; I want to know which option you find preferable.

The final for the course will be given in a week, but you will find a video describing the final and the grounds rules for taking it this week. The only consideration for earning a certificate of completion for FEE is a passing grade on the final. The “finals period” starts in a week and lasts two weeks. During that two-week period, you will have 48 hours to work on the final and submit your answers (three submissions are allowed). The 48-hour working period cannot be broken up. From the time you access the final's questions (and print them if you want), you must submit your last set of answers within 48 hours. As the video states, the final is what we term in the States an “open-book” final. You can look at any materials related to the course while you work the problems.

Sun 7 Apr 2013 4:24 PM CEST

This and next week's lectures concern information theory, a framework for data communication and signal processing developed by Claude Shannon in his 1948 paper. We will learn about lossless and lossy compression (LZW in the first category, mp3 in the second) and how data can be communicated over a noisy channel yet be received without error (error-correcting codes). These discoveries, all contained in his 1948 paper, set the stage for what is known as the information revolution. Digital systems are not prevalent just because they are different. We are going to show that their uses have a sound engineering justification; they can be better in many circumstances. I think these two weeks comprise the fun part of Fundamentals of Electrical Engineering!

Sun 31 Mar 2013 5:42 PM CEST

First of all, it appears the Coursera did not send out last week's announcement as an e-mail. You can see the Week 9 announcement on the FEE landing page.

This week we learn how communication systems work, both analog and digital. Analog here means amplitude modulation of a sinusoidal carrier (AM). Such radio systems may seem old hat, but it is very important to understand their capabilities. We then turn to digital systems follow, which send data (bits) from one place to another. Such data could be inherently digital, like text in SMS or Twitter, or the digital data could have originated from the analog-to-digital conversion of an analog signal. Such is the situation with digital television, modern cellular telephone, Skype, .... Both analog and digital systems combat channel attenuation and channel noise as much as is possible; turns out no other system can send a message through a noisy channel and retrieve it with a larger signal-to-noise ratio.

Sun 24 Mar 2013 4:02 PM CET

After eight weeks, you know the fundamentals of electrical engineering and basic circuits. Now we turn to technology for a while as we focus on communication systems. The fundamental model of communication introduced at the course's start comes back forcefully. In this week’s lectures, we concentrate on the channel: what Nature does to our signals as we try to communicate. I like to summarize a channel's behavior as “nothing good happens in a channel.” At least noise is added and the signal is attenuated (and it could be worse!), meaning the communication schemes need to be devised to cope with the channel.

Communication schemes fall into two broad classes: wireless (something with which you are very familiar) and wireline. It is important for electrical engineers to understand the pros and cons of each when making design decisions. For example, wireless channels are noisy while wireline channels are virtually noise free; wireless channels are very convenient (can be used almost everywhere) while wireline channels mean you are connected by a physical wire to the receiver. Welcome to the world of engineering design: lots of decisions to make.

Sun 17 Mar 2013 4:07 PM CET

How can choosing the way to implement digital filters be interesting? This week we discover two ways (can you guess what they are?): time-domain and frequency domain. When we discussed circuits, the implementation was the circuit itself. We used the frequency domain to better understand the input's structure and what the filter did to it. In digital signal processing, the computer implements the filter, but what algorithm does the software use? Several choices can do the job; the one we choose is based (almost) entirely on which runs faster. That goal often leads to a surprising design choice. You will likely be surprised what the signal processing software is doing in your smart phone.

Sun 10 Mar 2013 4:29 PM CET

This week, we delve into the heart of digital signal processing: computing spectra. Nothing new in defining the Fourier transform, but doors to new venues open when we consider how to compute the spectra of data. You will learn of the fast Fourier transform (FFT) that efficiently computes spectra. The FFT is so efficient that it enables many signal processing ideas, including the computation of spectrograms. Its re-discovery in 1965 caused an explosion of ideas about using digital systems rather than analog ones for processing signals. You will learn that there is a sound technical justification for choosing digital implementations in many applications.

Sun 3 Mar 2013 3:16 PM CET

This week marks another turn in Fundamentals of Electrical Engineering as we pass from analog processing of signals to digital processing. Measured signals---audio, pictures and video---must somehow be read into a computer for processing and subsequent transmission over computer networks. Can the transformation from analog to digital be accomplished without error? As we shall learn this week, the answer is complicated: yes and no. The last lecture begins the discussion of digital signal processing (DSP): the filtering and manipulation of discrete-time signals inside a computer, be it your laptop or your smart phone. We will go over familiar ground---complex exponential signals and linear systems---and learn about new issues in the DSP world.

Sun 24 Feb 2013 5:51 PM CET

This week, we reveal the entire picture that is the frequency-domain view of signals. Everything we have been doing comes into play now and, once all is revealed, you will know how to find the output of any linear, time-invariant system to any input.

You have already encountered the role of modeling seemingly unrelated phenomena with linear systems, even circuits. I am thinking of the "Long Days, Cool Nights" problem. This week, we will learn how linear systems theory applies to human speech production. This model is very important and is used to advantage in many circumstances. Because we have a manageable description of the speech signal (rather than one based on detailed physics), systems can be developed that exploit the speech signal's structure. I am think here of every cell phone. Did you know that the raw speech signal is not transmitted over cell phone networks? Instead, the parameters of a model of your speech are sent! We will talk more about how this is done later in the course. For now, we need to understand the speech model.

Who said modern electrical engineering isn't cool, even useful!

Sun 17 Feb 2013 4:23 PM CET

Well, you now have all the knowledge of circuits that you need for this course! You know how to analyze and build simple filters. Starting this week, we turn toward understanding what kind of filters (and other types of signal processing) we need for applications. The key idea we develop this week and next is thinking of signals in the frequency domain as well as in the time domain. It's very revealing to discover the different ways of thinking about signals, and then choosing the most effective one according to the application. We start with Fourier series.

A big surprise in the homework this week is using circuits to model basic features of the Earth's environment. You will derive a model for the earth's yearly temperature variations by analyzing real data. The frequency domain turns out to be the best way to analyze the data. Past students have found this problem to be really interesting: EE and the environment! Toward this end, you will need to become acquainted with Matlab or its public domain variant Octave. I made a video introducing these signal processing programming environments. Links associated with this lecture take you to a Matlab introduction and to the Octave download site (Octave is free but Matlab is not).

Sun 10 Feb 2013 4:12 PM CET

This week's lectures show how to solve any circuit, making it one of the most important weeks for this course. Everything we have learned about resistor-and-source circuits will be generalized in important, fundamental ways. Special this week are "extra" lectures. I had requests for material on op-amps (operational amplifiers), so I put up the videos. This material is not covered in the problem set, making these lectures truly "extra."

Later this week, I will be adding exercises to provide more examples. If you have not tried the exercises for the previous weeks, they provide a review. Exercises are graded as right or wrong but no numerical grade is assigned.

I have increased the numbers of attempts on homework to 10. I realize that many answers are graded as being wrong simply because of typing errors and automatic grader problems. Coursera will be changing the problem system to allow you to enter answers only for the questions you answer incorrectly. This change won't happen soon; in the meantime, I suggest you work the problems (they can be printed of course) offline and enter your answers in a text file. You can then copy-and-paste your answers. Rework/retype the incorrect answers and re-enter them; you can paste the answers to the questions you answered correctly from your text file.

Sun 3 Feb 2013 4:24 PM CET

The second week of lectures on Circuit Fundamentals (finally, voltage and current!) and Problem Set II will become available Monday.

Starting with this problem set, you will only be given a total of three attempts at working the problem. Each time you submit your answers, each of your answers will be graded as right or wrong, but explanations and answers to the problem set will be available only after the due date. You can rework the problems you missed, but you will need to resubmit answers for all problems. Since the homework problems will become gradually more challenging as the course proceeds, we highly recommend you to start the habit of printing out the problems (available through a link in the problem set preamble) and working on them with paper and pencil. Also, please be sure to read the problem statements carefully and double check your expressions before you submit.

Sun 27 Jan 2013 4:20 PM CET

In response to "popular" demand, I have just added a set of exercises (ungraded homework) that concerns how to decompose signals as a superposition of scaled and delayed unit-step and ramp signals. They are fairly easy and meant to help you understand superposition rather than test your knowledge.

Fri 25 Jan 2013 11:52 PM CET

For those wanting a complete copy of the notes in pdf format, you can download them here (3MB).

Tue 22 Jan 2013 7:30 AM CET

Many students have looked at the course's book online and can't wait to get started. I have set the first lecture, a course introduction more detailed than the video you have seen, to become live at midnight my time (UTC -600) Sunday. The rest of the first week's lectures as well as the first week's problem set (labeled as a Quiz in Coursera) will be visible on schedule a day later.

In the meantime, do look at the course logistics page, especially about entering mathematical expressions as answers (we'll do a lot of that!). Forum moderators are Rice students who have taken the course; they (I will help too!) will help figure out how to enter expressions as answers and answer questions about the lectures and the problem sets.

"See" you in a video soon!
Don H. Johnson

Sat 19 Jan 2013 12:00 AM CET

Fundamentals of Electrical Engineering will (finally) launch on 21 January. This course has been taught at Rice University for several years as the introductory course for electrical engineering majors. It is structured somewhat differently than introductory electrical engineering courses taught at other universities. Instead of concentrating on circuits and practical systems, this course goes over virtually all topics in the signals-and-systems subfield, arguably the heart of electrical engineering. Once you complete this course, you will know the fundamentals to some depth in the field and understand the principles behind the design of cellular telephone, Wi-Fi and computer networks.

I been asked by several students if the prerequisite of single-variable calculus suffices. That and a knowledge of complex numbers are all that is required. Reviewing complex numbers, addition/subtraction, multiplication/division and Cartesian/polar forms of representation would be a good idea. We really use complex numbers in electrical engineering a lot.

The text for the course is online (http://cnx.org/content/col10040) and was written for this course. Each video has links to the appropriate online modules so that you can read the lecture material and, if you chose, work short exercises to test your understanding. Additional homework problems to those assigned in the course can be found at the end of each chapter.

Can't wait to get started!

Don H. Johnson

Thu 10 Jan 2013 2:24 AM CET