Concerning "The Art of Electronics" as an entry-level recommendation

[slateraptor]

So I was in the middle of having lunch when my mind arbitrarily started to think about why anyone would recommend "The Art of Electronics" as a great starting point for someone with interest but little to no formal background on the subject. There are seemingly countless recommendations swearing by this book, but sometimes I have to question whether these people have actually read it.

Take, for example, the section in chapter 1 on Thevenin equivalence, summarily encapsulated on a single page as "Easy!" with Norton equivalent "incidentally" mentioned as an aside. Then right into a short description of specialty Zener and tunnel diodes. Then waveforms. Then the time-domain sol'n to an RC circuit. Then a discussion on impedance whilst brushing off Fourier analysis as "seldom necessary". Really? I mean this is just chapter 1 and it's already all over the place!

Flipping to an arbitrary page: 276 on state-variable filters, a whooping 2 paragraphs written like an advertisement for National and Burr-Brown; shimmy to the index and you'll discover that the only other mention of state-variables is on pg 295: another Burr-Brown advertisement. Flip two pages to Figure 5.24, which I instantly recognized as the topology of a synthetic inductor; with a single paragraph, no derivation, and not even so much as a precaution on its limitations, I couldn't help but wonder how many hobbyists over the years have blown up their circuits attempting to use this guy as a replacement for an inductor in a power circuit.

The structure and contents of the book makes sense after a brief glance through the bibliography section: the majority of references were handbooks. If the book weren't so bulky (and expensive), it would make a great shitter reader for a seasoned engineer. But as a point of entry, I'd be surprised if anyone without a formal background was able to walking away from this book with anything deeper than exposure to the lingo, if you will.

[Conrad Hoffman]

My opinion is a bit higher, but I've always wondered how good an entry book it is. If you're reasonably well advanced, there's little in it that's useful. If you're a raw beginner, it's missing quite a lot you need to know. It seems good for those with some introductory background, but who aren't yet familiar with all the hands-on basics you pick up the first few years. I know everybody dislikes math because its scary, but I think somebody is way better off with one of the introductory circuit analysis books like Boylestead (sp?). Usually they can be found cheap as used textbooks.

[gregariz]

But as a point of entry, I'd be surprised if anyone without a formal background was able to walking away from this book with anything deeper than exposure to the lingo, if you will.

You'll be attacked for that post because there are alot of fans of that book. However I completely agree with you.

For some funny reason its a topic that seems to come up for discussion every couple of years. The only conclusion I can come up with is that its because the Authors were with Harvard University. If they had been with the University of Nebraska it would have sunk without a trace a long time ago. I've got a stack of books that I think are all far better than that one.

My take however is slightly different to you. As I've worked with a lot of physicists and chemists my belief is that they don't understand the word synthesis or design process as an engineer is formally trained to do. As a result everything they do has an air of experiment about it. That's at least my take on why the book seems so incoherent - it doesnt flow because the authors were never taught that you need to learn things like norton's before you need to do an impedance match. In my travels it seems to be very popular with non engineers who have come into electronics I think for the same reasons - it gets you to a topic you might need without giving you some background that would appear in design orientated books.

[McMonster]

I'm a beginner and I got my hands on this book when I was starting. It all depends on how do you read this book, there's a lot of things that look simply scary, but what's great about this book is that you don't have to really pay attentntion to this. If I saw something I did not really liked I just skipped it and read ahead. Sometimes I suddenly got enlightened, went back, read it again and so on. Reading it word-after-word and trying to get it all at once is a mistake in most cases.

If you don't push it, it's a great book. And it's certainly better than all this crappy, boring, academic technobabble (any Star Trek fans?) handbooks, it's written with a much simpler language.

[IanB]

Maybe there's a reason it's called The Art of Electronics, rather than The Science of Electronics?

(Note: I've never seen the book, so I have no clue about the content.)

[robrenz]

I think this book is the best all around electronics book   http://www.amazon.com/Practical-Electronics-Inventors-Paul-Scherz/dp/0071452818/ref=sr_1_1?s=books&ie=UTF8&qid=1319339386&sr=1"-1  Practical electronics for inventors by Paul Scherz.
It does have some obvious mistakes in some of the figures but it is loaded with extremely usefull info I have not seen anywhere else. I consider this the most usefull electronics book I own.

[EEVblog]

I think of it as the best electronics reference book around, not necessarily the best one to actually learn electronics from scratch.
It's the one I pick up when I have to check something I haven't done in a long time.

Dave.

[slateraptor]

I know everybody dislikes math...

Speak for yourself. Why is it that math seems to be the only thing on this planet that society has given its blessings to suck at? Food for thought.

You'll be attacked for that post because there are alot of fans of that book. However I completely agree with you.

Alas, bring forth discourse, for it does not exist without dissenting opinion.

[Conrad Hoffman]

Well, every time I suggest  a book with good fundamentals that covers the math, it goes over like a lead balloon. Admittedly with LTSpice and Excel one can accomplish a lot, but electronics really should be a math-heavy discipline. I appreciate math. I even enjoy math. I use math. Unfortunately, I'm just not as good at it as I wish I was. I think it has something to do with memory and attention spa... look, there goes a squirrel! What was I saying?

[johnboxall]

A good contemporary book for outright beginners is "Electronics - A First Course" (3rd ed.) by Owen Bishop. See http://amzn.to/rRIXXT

Very approachable, the math isn't too bad however the basics are there and the explanations are excellent. Plenty of questions with answers as well. A great prelude to higher education or more complex books.

A note to my detractors: mathematics is necessary and interesting. When I recommend a book for beginners with some maths in it doesn't mean I am averse to it or don't recommend learning any of it. People need to start somewhere, and the Bishop book is a good point. Once a student has worked their way through it, they are more than ready for heavier texts such as AOE and so on.

[Bored@Work]

Ah, the Generation Arduino at work. No Math please, we are hip, and nothing that requires thinking because thinking sucks.

[alm]

Well, every time I suggest  a book with good fundamentals that covers the math, it goes over like a lead balloon. Admittedly with LTSpice and Excel one can accomplish a lot, but electronics really should be a math-heavy discipline.

I think what people like about AoE is that it covers the practical side that most EE texts lack, like what capacitor to pick (or at least which one you would pick twenty years ago). It goes much deeper than most beginner level texts. For example, someone recently referred to this text. It explains the transistor as a current amplifier and how to drive it (almost) in saturation, that's it. You won't understand how to build a current source (do you need another current source to feed it a constant base current?). AoE goes much deeper and shows all kind of applications to confuse and scare all beginners . Someone who has read AoE at least won't be completely out of touch with reality. But I agree that for people wanting to get farther into electronics, a more college level text with some proper math is a good read. I think the first chapter is probably the worst offender, going at a very high pace through all the passive circuit and miscellaneous stuff (plus diodes and power supplies) and give a bunch of practical tips in between.

[vk6zgo]

Ah, the Generation Arduino at work. No Math please, we are hip, and nothing that requires thinking because thinking sucks.

Well said,Oh great & grumpy one!

VK6ZGO

[urbanwriter]

Well, if 'electronics' should be a math-heavy discipline, why not say the same for people aspiring to be machinists? Let's see; geometry, trigonometry, algebra, arguably calculus and I'm sure I've missed strength of materials, metallurgy, machine design, optimization of design as well as dozens of other areas of study. Why, if you take the 'math heavy' requirement far enough no would ever have to plug anything in, or turn anything on, because they'd be studying the math that describes in intimate detail the physics of those operations. I'm amazed that given the dearth of math (me in particular, trust me) that anyone can do so much as drill a hole, let alone string a bunch of components together and still have the LEDs (plural, because in this case it is not driven by an Arduino) blink.

I will suggest that 'math heavy' may be appropriate if you're engineering for money - but then so is accounting, as a financial practice, for the absolute lowest cost for a given level of performance.

Part of the trouble is audience; I once got as far as LCR circuits in technical school, and that's when I bailed. Now, I'm coming back to electronics as a hobby, and trying to get 'enough' math to avoid the magic smoke. But I'm finding it difficult to find circuits (and the related math) that work together - in the way let's say that Dave's piece on heatsinks did - to tie math theory to the real world of smoking components. Sometimes the math/component/circuit problem is easy to arrange. Think of some number of resistors. Measure those resistors as accurately as you can. Now wire them in series. Now measure the voltage across the entire string of resistors. Now calculate the current. And now, just to live dangerously, splice your ammeter in to the circuit. How close is the measured current to the calculated current? Now that wasn't so hard was it.

OK. Now try to explain (calculus I suppose, but I'm only guessing, and with trepidation) what my scope is showing me when I look at the emitter in an oscillator circuit, that only has to drive a few LEDs. I suppose that for some people the mathematical representations that describe the rise, fall, and ringing that are all very evident on the scope don't need the scope. Those people might actually enjoy math, as a good friend of mine does, irrespective of its impenetrability for some of us.

And before I go back to smoking transistors. It is interesting to listen to people curse out (politely of course) all the applications where Arduinos (for example) have been used - too much complexity, too much wasted capability, 'why in my day,' - and even from my vantage point I often agree. But most of what most of us do with electronics as a hobby could just be purchased. A simpler, more elegant, and often more efficient in terms of cost/benefit analysis way to blink an LED. But should we just buy it?

Hopefully no one reads this as anything but a request for bestowing me with a more mathematical bent, or schematics that actually include test points, voltages, waveforms (heaven forbid anyone should duplicate Heathkit of days gone by) for troubleshooting. Or, indeed, something one or two steps higher in the pedagogical firmament than 'Make: Electronics.'

From Canada's rainy Pacific South-West, your math-disabled correspondent. 

[amspire]

A lot of the calculations we do are rough calculations and are probably no more complex then calculations about money.

Lots of people claim they are terrible at mathematics, but give them an hour to spend $500 and all of a sudden they are mathematical geniuses.

It is all about getting a familiarity with the way devices work so that the rough estimates of the component values you need are as easy as budgeting to buy presents for Christmas. In fact I think Christmas shopping is way more complex.

If you have a divider chain of resistors for example, you know the current through each resistor has to be the same so you don't even bother to calculate it. The voltage across each resistor will just be proportional to the resistor value.

If you have a transistor in linear mode, you start off saying to yourself that the base-emitter voltage has to be about 0.6V and the base current will be 100 or more times lower then the collector current.  Just from those two facts, you can design a transistor amplifier that will work successfully. You can usually work out how a transistor circuit you are looking at will work. No need to touch databooks, or find equations for the behavior of a transistor.

At minimum you have to understand Ohms law, but is that hard?  Most people probably understand that buying 1kG of beans for $5 is more economical to buying ten 100gram packs for $1. If you swap resistance for the unit price, current for the quantity and voltage for the total price, then Ohm's Law is exactly the same calculation. The only problem left is getting the same familiarity to the components that you do to money and shopping.

If I wanted to buy 10 cans of paint at $20 each, and I said I couldn't remember if I should multiply $20 by 10 or divide $20 by 10 to get the total price, you would look at me as if I was an idiot.  You didn't even need a formula to work it out - it just seems totally obvious.

Well with familiarity, you will know you have to multiply resistance by current to get the voltage - it is just obvious. Anything else doesn't make sense.

So if you are pretty good at handling money, then you are definitely capable of understanding entry level electronics design.

Richard.

[gregariz]

Well, if 'electronics' should be a math-heavy discipline, why not say the same for people aspiring to be machinists?

Hopefully no one reads this as anything but a request for bestowing me with a more mathematical bent, or schematics that actually include test points, voltages, waveforms (heaven forbid anyone should duplicate Heathkit of days gone by) for troubleshooting. Or, indeed, something one or two steps higher in the pedagogical firmament than 'Make: Electronics.'

From Canada's rainy Pacific South-West, your math-disabled correspondent. 

Sometimes I think one could get confused as to what current 'engineering is' or why you should subject yourself to sitting in a college math class.

My take is simple. That is, plainly, I can't get a job worth having (ie pays) that involves playing with test equipment, fidling with discrete transistor amplifiers, programming an Arduino or even laying out a PCB.

So what pays? Well in my current job we design software defined radio's that crunch digital satellite signals 30 channels wide all ripping into an Arm 11 at 55MHz. In order to do this we need to design our own Analog RF Asic's and a custom DSP Asic's. We are a team of about 8 people. We also had to write the decoding algorithms from scratch. - Yes there is alot of math in it.

In my first job I was hired to write mathematical algorithms to synthesize field patterns - again alot of math.

So these are the types of technologies that keep tech businesses in decent curry these days.

Now you can find 'design jobs' that don't include math and involve an 8 bit pic or similiar but these are probably less interesting, less well paid and probably less long lasting as well. The 'easy' no math stuff has been commoditised.

[Wim_L]

Probably the same phenomenon at work as when Feynman's lectures are recommended as good introductions to physics. They do cover introductory physics very well indeed --- For someone who already knows introductory physics.

As a true first encounter with the material, they're not quite as good. Introductory books are hard to compare anyway... By the time you get to the second one, you're not a beginner anymore and may miss things that would be stumbling blocks for a true beginner.


The value of AoE lies mostly in being different from the majority. There's no lack of textbooks that will teach you the equations. Real schematics, with real components, and advice on, say, which type of capacitor to use in what situation is somewhat less common in introductory texts.

One that might be good is "A Practical Introduction to Electronic Circuits" by Martin Hartley Jones. A bit smaller and more approachable, covers far less than AoE, but still with real components and recommendations to build and test the circuits presented.

[Alex]

We are a team of about 8 people.

Gregariz, I would be grateful if you could share how big is the company and how exactly your team fits in the organisation.

[robrenz]

I mentioned "Practical electronics for inventors by Paul Scherz" earlier in this thread.  I have the Art of electronics also. Practical electronics for inventors is hands down a better book for a all encompassing view of electronics and electrical.  I think the title scares people away in that it would not be technical enough. This book combines 260 pages of in depth theory (which he tells you can be ignored if you want to) with an additional 674 pages of extremely practical electronics. It is presented in a style that is very accessible with lots of graphic support.  I have attached two example pages of component identification and tolerance and the table of contents.  Remaining pages will be on next post.

[robrenz]

Remaining table of content pages

[alm]

It seems that the second edition of Practical electronics for inventors was significantly expanded compared to the first edition, since that was only about 600 pages. From a quick glance, the chapter on passive circuits appears to be more extensive than AoE, but the transistor circuit covers less ground. I didn't find a mention of the exponential relation between the V_BE and I_B (Ebers-Moll), and it also spends less time about how h_FE is not a very stable parameter. Without these facts, it's hard to understand correct biasing and amplifier distortion, although it does mention r_e. Maybe it's fixed in the second edition, I don't have access to it. Would this topic have been answered after reading that book, especially the later questions?

[gregariz]

We are a team of about 8 people.

Gregariz, I would be grateful if you could share how big is the company and how exactly your team fits in the organisation.

The company is now around 250 people, alot of them sales spread around the world. Engineering is clearly split in two - R&D and Applications. We are probably getting a bit big now - as a general rule I have come to believe you really want to be in the smaller ones (although probably not too small unless its yours). Not sure what I will do after this - might try my own for a bit.

[robrenz]

It seems that the second edition of Practical electronics for inventors was significantly expanded compared to the first edition, since that was only about 600 pages. From a quick glance, the chapter on passive circuits appears to be more extensive than AoE, but the transistor circuit covers less ground. I didn't find a mention of the exponential relation between the V_BE and I_B (Ebers-Moll), and it also spends less time about how h_FE is not a very stable parameter. Without these facts, it's hard to understand correct biasing and amplifier distortion, although it does mention r_e. Maybe it's fixed in the second edition, I don't have access to it. Would this topic have been answered after reading that book, especially the later questions?

The Art of electronics is definitely more in depth in transistor theory and many other areas.  It is a great book and much more useful to the more experienced.  It would cover that topic in much more detail than Practical electronics for inventors.  but it also does not cover a huge amount of information that an entry level person needs to know.  I am not bashing The Art of electronics, everyone serious about electronics should have it also.

My point is that this topic is "entry-level recommendation" and in my opinion Practical electronics for inventors covers everything a beginner would need to be aware of in one book.  As an example: 240/110house wiring what does common and hot mean, earth ground ?  what does that have to do with working on AC powered equipment.  What do these components look like?  what do the colors and numbers mean?  what type of capacitor or resistor etc. should I use for this situation and why?

PS. I am far from entry level but I am also very far from experienced, so maybe that is why I think Practical electronics for inventors is so essential.

[HLA-27b]

One book that definiteley deserves a palce here is

Fundamentals of Analog Circuits
Thomas L. Floyd
David Buchla

http://www.amazon.com/Fundamentals-Analog-Circuits-Thomas-Floyd/dp/0130606197

Before reading this book I was quite miserable at electronics. I used to look at schematics rarely recognizing anything except the most rudimentary. When reading the texts I carefully skipped the maths bits always ending up only with the most superficial and anecdotal out of it.

Fundamentals of Analog Circuits lets you penetrate the topic with nice explanations and graphics like any good book but the difference is that it lets you get a grip of the maths too. It assumes very little in terms of math background and works step by step until building a good picture of the topic. The transistor topc attached below is a good example.

Very highly recommended to anybody who wants to go past "noob" in electronics.

[Alex]

We are a team of about 8 people.

Gregariz, I would be grateful if you could share how big is the company and how exactly your team fits in the organisation.

The company is now around 250 people, alot of them sales spread around the world. Engineering is clearly split in two - R&D and Applications. We are probably getting a bit big now - as a general rule I have come to believe you really want to be in the smaller ones (although probably not too small unless its yours). Not sure what I will do after this - might try my own for a bit.

Thanks gregariz, useful. Good luck with any ventures.