A learner's rant

[David Aurora]

I've been teaching myself about electronics over the last few years from various books, websites, experimenting and conversations with smarter people than I. With every source of knowledge (besides experimenting), I have the same complaint, and it's reached the point lately of driving me fucking berserk (and I've realised it's why I've abandoned so many attempts to learn in the past), so I'd like to humbly request right here and right now that nobody producing educational content ever again uses the following words together:

"Let's assume that..."

No. Let's NOT assume anything. Tell me why it's the case. Whether it's choosing a collector current, or a filter cap value, pull up/down resistor values, input/output impedances, fuse sizes, whatever, please tell me WHY we're using that value/voltage/current/etc. Don't shortcut, teach it properly or don't bother.

I understand that at a certain point you collect a certain amount of knowledge and can fill in the blanks for yourself. I'm reaching that point myself now in lots of situations. But it has been hell to get there. It seems like every damn book or tutorial I ever find on a subject starts on some unexplained assumption and leaves me scouring 20 other books or sites for an explanation of why we started with that assumption. I have so many dog eared half finished books laying around because the information is completely useless to me until I find out why they chose certain starting points in a circuit. The info is fine to copy their design, but tells you absolutely stuff all about designing your own circuits for your own requirements.

I hope this rant makes sense. It's hard to explain exactly what I mean. Argh. My brain is fried right now. I just got halfway through reading another whopping explanation on a topic only to find another value chosen to explain equations with no explanation of why out of every possible value available you'd choose that one. Is it related to impedance? Power dissipation? Stability? Frequency response? NFI.

Now... off to go figure that out from another 50 sources before I can continue 


P.S. This is why I'd love to see Dave tackle some super basic beginner concepts at some point, he makes things so much easier to understand. Maybe it's because we both speak Strayan, I dunno

[Bored@Work]

And you know what would happen if your wish was granted? You would be coming here and complain why they use pages after pages full with Math to explain details, while never getting to the point.

[Psi]

This issue is a bit complicated, because everyone learns in a different way.
However my perspective is...

You need some level of "lets assume that.." otherwise things are too complicated to get your head around.
It's the same with computer/mcu programming, you need abstraction to break things into chunks you can understand.

A lot of people make the mistake of trying to understand everything (or believing they should know everything). It just leads to getting depressed and annoyed.
Even the most experienced engineers don't know it all.


It's a lot less about "what you know" and more about "what you can build"
Also its about having the skills to research areas you don't understand in order to get something to work (even if you don't really understand why it works).

Everything you build teaches you something and that enables you to build more complicated things in the future.

[David Aurora]

Bored@Work- That's a bit straw man, really. There's no need for it to be taken to ludicrous extremes, just simple one line statements here and there. As a super simple example, say you're explaining voltage dividers. You can give the formula and an example quite easily, but it doesn't really help a beginner much staring at a pile of resistors going "Well should I use the 100k resistors or 10k resistors". Mention the (obvious to us at this point) current draw aspect, and they'll have a much better idea of how to select suitable values.

Psi- Yeah I hear what you're saying, I just think there's better ways to explain things. To put it into something I've got more expertise in than electronics, let's jump to music. When I was taught chords by various people, they generally went "Put your fingers here, here and here. That's a C major chord". Whereas the way I've taught people is "The major chord is the first, third and 5th note of the major scale". The former explanation lets you repeat the chord you've been shown, as it was shown to you. The latter let's you find any major chord anywhere you want on any instrument you want. So yeah, you definitely need a certain level of knowledge (here it's knowing the major scale). BUT... this step is a complete lesson built off that prior knowledge, it doesn't gloss over anything the way the copy/paste "Put your fingers here" method does.

[EEVblog]

Working at a French company once, I decided to take the French language class they offered.
I lasted a couple of weeks before I quit.
They started right at the frigg'n beginning with all the technical aspects of the language, with all the accents and forms etc 
All I wanted to do was learn some basic phrases etc, we never got to that part 
It was like trying to learn electronics by starting with quantum theory and maxwell's equations.

With the original question, it's ok to say "Let's assume that..." when the value doesn't really matter much and you are trying to teach a higher level concept.
Getting bogged down in the details can be boring.

I actually have on my to-do list a video just about pull-up resistors. Because someone asked that once on here (you?), and I was curious to see how long a video I could make on the details of choosing one of the most mundane component values in electronics

Dave.

[ftransform]

they need to fix firefox so it does not shit itself when I have 100 tabs open, 100 tabs solving the problem you indicate.

I do not bother with paper textbooks. Open two textbooks on the same subject, 3 app notes and 4 web pages to get a good understanding.

[Mechatrommer]

lets learn opamp subject!... a simple one, voltage follower... "lets assume that"... no bias current no voltage offset no temp drift super linear opamp bla bla, so we get the equation.... Vout = Vin+, simple! fun! now lets dont assume the above mentioned are zero, try to put those into the equation and see? i dont thing thats fun esp at introductory chapter.

[jeroen74]

I think 'The Art of Electronics' has a good approach. Start out with the ideal component assumption, and later on talk about the imperfections of real components.

When I was taught chords by various people, they generally went "Put your fingers here, here and here. That's a C major chord". Whereas the way I've taught people is "The major chord is the first, third and 5th note of the major scale". The former explanation lets you repeat the chord you've been shown, as it was shown to you. The latter let's you find any major chord anywhere you want on any instrument you want

That's the subtle difference between training and teaching I guess.

Like the Microsoft Word trainer who regards Ctrl+S as the wrong answer to the exam question: "How do you save the document?"

[jucole]

I find the best way to learn is to design and build a project that includes lots of things you want to learn; then just build it!   Soon enough you'll run into lots of weird  phenomena that makes no sense or if you're really unlucky and your project works first time; great!  but if you're lucky enough it doesn't actually work;  that is when you learn the most! ;-)

[ptricks]

Use some of the online video courses from the major colleges. Like anything though it is going to depend on the teacher. Some teachers are great while others shouldn't have ever been teaching.  I learn best by doing, someone just talking about things makes me start to daydream in class.
Free video tutorials.
http://www.infocobuild.com/education/audio-video-courses/electronics/electronics-and-electrical-engineering.html

[grumpydoc]

They started right at the frigg'n beginning with all the technical aspects of the language, with all the accents and forms etc 

That's the French for you, unfortunately.

[G7PSK]

The best way to learn any language is to go to the country where that is spoken, the same principal applies to electronics go where it is spoken and mix with expert speakers and ask questions and get down and dirty and do.

[AndyC_772]

This is something I found highly frustrating many times, both at school and at university. Many people who are experts in their field really have no concept at all of just how difficult their subject is for someone who has no pre-existing familiarity with it. Often it's as if they start teaching at chapter 2 of the course text, and leave out those crucial first pages entirely and without even noticing it.

Earlier this year I started writing a book. It's a beginner's guide to electronics, designed for the curious hobbyist who wants to learn about the subject in a practical way, without getting bogged down in maths. Most importantly of all, I wanted it to give the reader a really solid, intuitive feel for what each component really does, and how its presence might influence those components around it.

The way I approached the subject was by analogy: electricity is like beer. We can all imagine beer, and it captures interest in a way that abstract concepts and equations never will.

Sadly real work got in the way, so I only got as far as the first couple of chapters before having to shelve the project. Maybe I'll pick it up again if work goes quiet again next year. In the meantime, if enough people ask, I might be persuaded to post a chapter...

[Mr Smiley]

The problem with assumptions is there based on what the person making the assumption already knows. Their brain flashes through all their past knowledge and experience in a moment and up pops a value. They won't be able to give you a 3 seconds explanation because sub consciously it's been done for them and there told from inside that it's gonna be X.

This really is a major problem, particularly in books.

When i was taught programming with pascal and machine code, electronics and all other things i wanted to understand i was in a fortunate time.

Books in them days, the 70's and 80's described principles and examples in simple to understand ways, say for example programming C and pascal. If you look at today’s books you have a  few pages of 'for' 'while' and 'if', then the rest goes into brain deadening masses of unexplained dribble.

If you  get hold of the 'old' books, they give pages and pages of clearly described code full of comments and what you shouldn’t do and why.

Same with today’s electronics books, there written by people who just want to churn out what they know, get the 'old' books from the 50's and 60's, even valve based, and the detail and explanations is beyond belief.

A lot of the foundations and basics what you may want to learn is no longer available in today’s publications or on the internet, as you've discovered it's all assumed you know it all.  It has gone with those that knew and have since died. As an example, about 25 years ago i wanted to build my own studio flash units, at that time 'bowens' ruled supreme, you couldn’t by the tubes or the caps or anything to build these big beasts. There was no internet, and the library ruled. But the only info i could find, very little but very detailed was in books from the 50's. I found a small company that made the tubes, went and fobbed my way in and tried to make out i knew what i was talking about. From that i got the name of a designer, got his phone number and phoned him up,  The amount of info he was giving me just went on an on, after 10 minutes i was forgetting what he had started telling me, so i hung up, grabbed a tape recorder, phoned him back, apologized for being cut off and then recorded over an hour of information that to this day i have never heard again. Did you know that high power studio flash units can self trigger in complete darkness, illuminate the room with almost no ambient light and they never will.

Another thing you may find is that you will learn more from what you 'notice' than from what your told.

 

ps , get the old books, i have books from librarys that were put out of circulation and had their hard covers ripped from them do dissuade people from taking them from the bins, old, damaged but worth their weight in knowledge that is now disappearing.

[robrenz]

I use two modes of learning.

One is absorbing whatever I am capable of with my current knowledge base as I read the forum and watch videos. If something sparks a deeper interest I may spend some time researching it. I absorb what I can follow and I don't worry about what I can't follow at the moment. I usually gain some incremental amount of understanding. I am interested in all of it but I realize I am not going to be an expert or even adequate in all of it.   Over time my "knowledge + understanding = wisdom" grows. I am enjoying the journey, I have no specific destination.

Second is doing a project that provides a NEED to learn. I pick something that I am inspired to do because I know its going to get ugly and I will need staying power to not quit. So I decide to build a better/more usefull decade resistance box  to sell. Who can't handle a decade box design? So that has led to a 1-1/2 year obsessive study of resistance, resistance measuring etc. etc. With a many rabbit trails inspired by the forum of mode one learning thrown in. I don't just read, I DO/BUILD things that reinforce the concepts and burn it into my brain. So now I know a lot about resistance and measuring it, I may not be an expert, but I can explain all the important concepts to someone off the top of my head. I will never forget this knowledge because I have used it long and hard enough that it is relatively permanent. My overall electronic knowledge/wisdom has increased in the process. On the other hand I could not build a simple transistor circuit without looking up EVERYTHING because I don't have any experience or a current need.  I know the basics but not the details and that doesn't bother me.  

So now onto the next project, making a controller for my 4 quadrant PS so it can be an Eload

[poodyp]

I personally love the Arduino "tutorials" that ammount to "plug this widget in and use this code. Put function_call(4) in your code somewhere and tada, you're totally a programmer dude".

[IanB]

Another point to remember on this subject is that the information may not all be in the same place, or under the same heading.

For instance with electronics, the basic assumptions of things that "everybody knows" may not be found in electronics textbooks at all. At some level you will have gone back to basic physics and the assumed information will be in physics textbooks. It may be things you might have learned in science classes at school under a quite different subject heading.

Similarly with advanced material like the discussion on the operation of studio flash units. This information might be consolidated in a highly specialized text on the design and manufacture of high power studio flash units if you could ever find such a text. However, all the information is most likely out there as quite general and un-specialized theory about the physics of gases and plasmas and gas discharges, if you knew how to interpret it and put it all together.

Experts become experts by learning all sorts of general and widely dispersed material, combining it with experiments, and salting it with time and experience. Experts themselves may be lost when they retire, but the information is not lost. It just drifts apart and scatters again like dust on the wind.

[free_electron]

I actually have on my to-do list a video just about pull-up resistors. Because someone asked that once on here (you?), and I was curious to see how long a video I could make on the details of choosing one of the most mundane component values in electronics

Dave.

my speculation is a good 4 hours ... don't forget topics such as adjusting for the line capacitance , differenc ebetween carbon  , metal film , power dissipation in the resistor, how to place via's on the board ,difference between series termination , parallel termination , source matching , load matching , daisy chain
. don;t forget resistor netorks. single pin common , multiple pin common.

here's a nice one... i got calle din to help debug a system that kept reboot endlessly. it was built around a 68000 processor. the power supply kept collapsing... the doofii ( plural of doofus .. ) had install 1k pullup resistors on all address, data, and control lines to make sure to have nice 'hard' levels

so whenever the cpu wanted to access memory location 0 all 32 address lines 16 data lines and some control lines all were driven low , effectively putting like
50 1k resistors parallel with each other ... thus loading the power supply with 20 ohm. poor 7805 was already heavily loaded by all the chips and threw a fit when that happened. thermal overload kicked in , board screwed up...

there such a thing as overdoing it.

oh, and don;t forget to mention active pull ips with edge accelerators. linear technologies makes that ( amongst others )

[free_electron]

on the 'assume that'

good luck tring to explain the classic astable mulitvibrator with two transitors.. to explain that you need to assume that one transistor will go in conduction faster than the other and this the system will begin oscillating.

as for why that one transistor reacted faster ... and why it is not necessarily always the same one that reacts fast we need to delve in transysteor physics , transistor construction and the laws of probability , statistics (and other lies) ... and may have to drag in schrodingers cat ...
The mere fact that you are watching the LED's blink has an impact on which of the two transsitors will conduct the first.

is that all necessary to explain a simply circuit made form 4 resistors, 2 transistors 2 caps and 2 led's ? i don;t think so. so 'assuming that' is perfectly fine.

[jeroen74]

Until the learner simulates the circuit in, say LTSpice, and finds it doesn't work

[free_electron]

Until the learner simulates the circuit in, say LTSpice, and finds it doesn't work

Take computer that has spice installed and make a pile with computer, installation media , manuals , whatever that relates to spice.
Douse pile in gasoline and set on fire. After fire self-extinguishes : collect remainders and grind them down to powder.
Make a mixture of cement , water and sand enough to fill a 1 cubic foot container. Add aforementioned ground-down remainders to mixture and pour in mold. Once the mixture has set : bury hardened concrete block 6 feet deep and plant an apple tree on top of it.
Affix sign to apple tree : not fit for human consumption.

What i'm trying to say : Beginners (and even moderately advanced)  should be kept away from things like spice at all costs until they have fried enough parts and genained enough prectical experience with real parts to understand the problems with simulations and their flawed models.
Experience in electronics is linearly propertional with the amount of fried circuits and test equipment.

[David Aurora]

OK cool, looks like I'm not crazy, people got what I was talking about 

On the points about learning by experience, yeah, that's always the main thing. I want to DO things, not just KNOW things. But you definitely need to start with a little bit of study, there's no way to instinctively know how to build a rectifier, or what an op amp does, or Ohm's law, any of it. You need the basic idea of how to do what you want as well as references now and then while learning.

Mr Smiley- The old books suggestion sounds like a good one. I picked up a couple of books on transistors from the 60's a while back, and they're much better than the newer explanations I've come across. Basic electronics knowledge is assumed (e.g. You need to know Ohm's law, what a resistor or capacitor is, how to connect things together etc.) but no prior knowledge is assumed for THE TOPIC THE BOOK IS SUPPOSED TO TEACH YOU ABOUT. Books like that aren't really there to show you how to build 10 different things, the point is to teach you the ins and outs of the component so by the end of it you can figure that out for yourself.

But I guess like the Arduino example, it comes down to what the readers want, instant flashing lights or a proper understanding.

[lewis]

Take computer that has spice installed and make a pile with computer, installation media , manuals , whatever that relates to spice.
Douse pile in gasoline and set on fire. After fire self-extinguishes : collect remainders and grind them down to powder.
Make a mixture of cement , water and sand enough to fill a 1 cubic foot container. Add aforementioned ground-down remainders to mixture and pour in mold. Once the mixture has set : bury hardened concrete block 6 feet deep and plant an apple tree on top of it.
Affix sign to apple tree : not fit for human consumption.

What he said!

"Let's assume that" is absolutely fine in my opinion, as long as the text immediately goes on to explain why the assumption is invariably wrong.

Autodidacts unite!

[vk6zgo]

on the 'assume that'

good luck tring to explain the classic astable mulitvibrator with two transitors.. to explain that you need to assume that one transistor will go in conduction faster than the other and this the system will begin oscillating.

as for why that one transistor reacted faster ... and why it is not necessarily always the same one that reacts fast we need to delve in transysteor physics , transistor construction and the laws of probability , statistics (and other lies) ... and may have to drag in schrodingers cat ...
The mere fact that you are watching the LED's blink has an impact on which of the two transsitors will conduct the first.

is that all necessary to explain a simply circuit made form 4 resistors, 2 transistors 2 caps and 2 led's ? i don;t think so. so 'assuming that' is perfectly fine.

This is exactly the example I was going to use!
Another problem,is if we don't either cover the whole gamut of human experience,or alternately,say "let's assume",& instead,just pick a transistor at random,saying say,"Q1 is initially saturated",you will have a huge number of newbies agonising on this forum about why some other book or website chooses Q2.

[miceuz]

With every source of knowledge (besides experimenting), I have the same complaint, and it's reached the point lately of driving me fucking berserk

I know your pain bro! I'm going the same path for the 3rd year, and I've come to some conclusions:

1. The majority of content online is crap. To get semi decent understanding of a concept you are digging into you have to read piles of semi complete tutorials and cherry pick facts to fill the gaps.

2. I should have gone to university to learn the subject. A good one university. But that does not count for me anymore as I'm too old and too broke for that. Online courses from MIT and Cornell are great, but when you can't ask the professor your own questions, it still leaves some gaps unfilled.

3. electronics.stackexchange.com and forums like this one is a great place to fill in the gaps. You have to do it yourself and it amounts to the same cherrypicking, but at least there is a way.

4. When you get some knowledge of the subject, compile it and post it online so generations to come will be able to find another resource with some gaps filled for them.

Just right now I'm compiling a post about transmission lines, termination, driving long lines with digital signals and how long is too long. Yesterday I just got "The art of electronics", I opened the page about "Digital signals and long wires" and man, how disappointed I was - "just put a resistor of value equal to characteristic impedance of the line in series with capacitor to ground. Something like 100pF should do ok" wha? I've expected a better explanation, like different termination techniques, pros, cons, harmonics, rise times, etc.. ok, back to cherrypicking