Learning Paths

[icharters]

I'm trying to figure out what the smartest path is to take regarding learning EE as a hobbyist right now.  I bought a ton of stuff of Aliexpress like resistors, caps, transistors, etc.  Probably $200 of assorted EE frequently used items just to get me started with some essentials.  I also bought a Siglent SDS 1140X-E scope, have a couple of bench power supplies, hot air, iron, etc.

I own Art of Electronics and was thinking about buying LtAoE, but getting the parts seems to be a bit of a screw around.  Some things are obsolete, a lot of the parts can be substituted for cheaper stuff apparently, etc.  Digikey doesn't seem to even sell it anymore - it has been on backorder forever.  Should I just bite the bullet and buy all the LtAoE stuff and find the substitutes for the things no longer sold? 
 
I saw that there are some MIT courses available as well, I was considering perhaps going that route and just starting from the beginning and working my way through the courses.

Any advice?

[Siwastaja]

Take things apart. Interested in power supplies, for example? Open broken power supplies, trace how the components are connected, try to see what topology they used.

Appnotes are quite good (sometimes they are crap, but academic books can be crap as well, so...). Hit any practical question, just Google it and read the appnotes from component manufacturers, discussing it using practical terms.

Finally, use a quick iterative design-build-debug-design-build-debug cycle. "Debug" in this case means learning, and includes reading on the subject. The field is so vast that without this process, you don't know where to start, and may be tricked into spending a lot of time in some completely theoretical part which ends up never paying back in practical knowledge. But if you let the practical circuits drive your theoretical learning process, it works out.

You did the right thing with basic components, scope, power supplies, hot air and iron. This is exactly the basic set you need. A good stereo microscope would be the next one, but it will blow your budget, comparable investment to the oscilloscope. (The fact is, modern extra-miniatyrized SMD IC's offer very handy features, for good prices. If you avoid them because you can't solder them, this limits your ability to do and learn things by quite lot. With hot air, soldering iron, and a good stereo microscope you can use almost any part on the market! You can do dead-bug prototypes with small QFN's, BGAs, and so on in mere hours.)

Get some unetched bare copper clad PCB material. It allows you to "deadbug" or "airwire" circuits on a top of a good ground plane. Very effective in testing out ideas.

I started getting into power supply design by driving a random MOSFET lying around, soldered to a piece of copper clad, driven from a function generator, connected to a transformer I wound using totally random pieces of ferrite and magnet wire. These projects proved successful proofs-of-concepts in mere hours, very rewarding. Such experiments are a vital part in the iterative learning cycle: read-experiment-read-experiment. If you try to do it with books and courses only, you never properly succeed, and it's slow anyway. If you only experiment, you can develop the whole theory you need, but it will take decades. But combine the two in an iterative process, and you learn quickly (and it's fun as well).

[garethw]

Just as Siwastaja wrote but in addition I would say the most useful thing I have found is to have a goal. There is so much stuff to learn about that it can seem daunting.
So some previous projects of mine have been a temperature/humidity display for my disabled son’s bedroom. This required me to learn how to program and code a PIC24, I2C interface with a Bosch BME680 sensor, layout a through-hole pcb with Eagle and prepare the Gerber files for manufacturing.
Recently I’ve started exploring surface mount components and have just received my first SMD pcbs, again, designed on Eagle. This was a breakout board for an STM32 mcu.

Clearly what you learn about has to interest you to keep you engaged.


Sent from my iPhone using Tapatalk

[MosherIV]

I found doing projects where I want something slightly different from examples I have reserached has helped me to to learn and understand how the circuits work.
I started by being interrested in linear audio electronics. Simple op-amp mixer circuits, guitar effects peddles.
Bench (linear) psu, lots to learn: emitter follower for the output stage, instrument op-amp for current limit
The point is, to design or modify a circuit you trully have to understand how the circuit works.

[techman-001]

I'm trying to figure out what the smartest path is to take regarding learning EE as a hobbyist right now.  I bought a ton of stuff of Aliexpress like resistors, caps, transistors, etc.  Probably $200 of assorted EE frequently used items just to get me started with some essentials.  I also bought a Siglent SDS 1140X-E scope, have a couple of bench power supplies, hot air, iron, etc.

I own Art of Electronics and was thinking about buying LtAoE, but getting the parts seems to be a bit of a screw around.  Some things are obsolete, a lot of the parts can be substituted for cheaper stuff apparently, etc.  Digikey doesn't seem to even sell it anymore - it has been on backorder forever.  Should I just bite the bullet and buy all the LtAoE stuff and find the substitutes for the things no longer sold? 
 
I saw that there are some MIT courses available as well, I was considering perhaps going that route and just starting from the beginning and working my way through the courses.

Any advice?

As mention/hinted here I think the best and only way is to start building gear.

The electronics world is far too big for you to become a expert in everything, so you must specialise in a particular area, one that interests you is best, one that pays dollars is also good.

You have the fields of  rf, audio, instrumentation, automotive, embedded, power, IoT, medical and many more to choose from, so pick one and start building the things that interest you. For instance a bench power supply is a worthy and vital project which should supply plenty of learning and building challenges for you.

My first project was a bench power supply built by following the plans in the 1954 ARRL handbook. A wood chassis, 400 volts dc via a type 80 valve rectifier, brass terminals, everything made from scraps as I had no money and electronic parts were scarce when I was 12 years old. That one project taught me a lot and lasted well into the "transistor era".

Once you start building, you will soon know what parts and equipment you need and making do with what you can get, instead of what you want, so as to achieve your aim, is to me the art of engineering.

Expect it to take decades, even a whole lifetime!

[Siwastaja]

I find that once you know the basics (that apply to all sub-areas), it takes about a year of hard work to become a semi-expert on something, and about a decade to become a real expert on something.

So you may eventually be able to pick about 5-10 "semi expertise" areas, and zero to one (or maybe two) where you can really shine (world class).

You really can't just sit down and "decide" what areas you are going to learn, you'll pick them up as you go, driven by your interests of that time, and actual job opportunities available.

It's also dangerous to be a one-trick pony.

Just to give you an example of what those "subareas" might be, I "semi expertise" in power conversion, motor controls, li-ion management, imaging & time-of-flight ranging, microcontrollers, embedded computing, and FPGAs. I have no idea whatsoever about RF, very little experience on high-speed signal integrity work, or high-precision analog measurement circuits.

Many real projects combine several areas of expertise, and those projects won't always end up being split to a large team of one trick experts because it may be more efficient to use someone who can handle a bigger part on their own, reducing communication overhead and risk. If you have a team of ten and everybody must succeed in their part, it's much more risky than if you just had three out of whom any two can complete the job.

[rstofer]

I have LTAoE and TAoE but, for getting started, I'm beginning to think that "Getting Started In Electronics" by Forrest Mims is an easier way to get going.

Other than kits of resistors and capacitors and MAYBE transistors, I wouldn't buy a bunch of parts.  Jameco has parts kits with cabinets.  I emphasize cabinets...

I don't think I am a fan of buying floor sweepings from AliExpress.  With Priority mail, I can get parts in 3 days or so from DigiKey.  I buy what I need plus extras.  I will also add something a little more pricey (Arduino or Pi or <something>) such that the order is worth processing.  But I buy for the project, not to build an inventory.  I get the inventory when I buy in multiples of 100 or whatever.  At $0.02/ea, there's no reason not to buy x100 resistors of several values on each order.

There are really two branches or a fork, if you will, on learning elecgtronics.  For the hobbyist, a few simple laws and a few equations will do the job.  They are mostly going to build stuff they find on the Internet and hope for the best.  Maybe they get augmented help here on EEVblog.  The other fork is electronics from an engineer's point of view.  Actually, this doesn't involve as much building as it does math and equations.  This fork is a lot of work and I can't say that it is much fun.  I would think that the MIT courses are going to be rigorous and most hobbyists aren't going to want to spend the time.  Look into "Real Analog" at Digilent for a shorter course but still covering a lot of material.  There are text documents, videos, exercises, lab experiments and it's taught at a level that more hobbyists will comprehend.  Unless you have had a lot of math, in which case, rock on with MIT!

I rather expect that most hobbyists want the adventure, not the explanation.

'No, no! The adventures first,' said the Gryphon in an impatient tone: 'explanations take such a dreadful time.'

― Lewis Carroll, Alice's Adventures in Wonderland

[IanB]

I'm trying to figure out what the smartest path is to take regarding learning EE as a hobbyist right now.
...
Any advice?

There are two approaches, the scientific and the utilitarian (or both).

The scientific approach is exploration. In this approach you think of gaps in your knowledge, where you can ask questions of the form "I wonder...?". I wonder how this thing works? I wonder what happens if I do this? I wonder if I can get something to behave in a certain way? (Detect motion, make a sound like a violin, respond when I touch it, ...?)

The utilitarian approach is to make useful things. Sound an alarm when your fridge/freezer is too warm, tell you when you accidentally left your garage door open, make a light that runs off nearly dead batteries, ...

You could pick either way, but you should definitely have some objectives in mind. Without a goal, your path will be aimless.

[Siwastaja]

I tend to always start something the utilitarian way, but get really interested in the process and end up with a more scientific approach. The utilitiarism then works as an additional source of motivation (at least it helps you fool motivate yourself into thinking that this can turn into money later and hence you can concentrate on it without worrying how to get the bread on your table...)

So it can be a mix of the two.

[icharters]

Thanks for the input, guys.  I think I will take the advice here and learn things on an as needed basis, and take a more scientific approach to learning it when i really want to deeply understand something I'm working on.
 
I think I'm going to build a power supply to start with and see where that takes me.

[rstofer]

Keep it simple, about+15/-15 and a separate +5 all rated about 1 amp will cover a lot of projects.  Search for 7815, 7915 and 7805 circuits.  Resist the temptation to overspec the PS to triple output 30A 3A.

[pwlps]

I think I'm going to build a power supply to start with and see where that takes me.

Oh, another one, since I joined this forum I see so many beginners pushed to start with a power supply project, seems we are really short of ideas here    For me a power supply is absolutely no fun, a top most boring stuff: can't play music nor make dance some remote stuff with it,  not even a way to have some fancy display, really nothing to play to the gallery (kids/wife etc.) 

[Jwillis]

Every circuit ,no matter how complex it may be, is just a lot of simple circuits put together .This includes IC 's .When you fully understand the simple circuits and how to identify them then and how they function ,the complex ones become much easier.
Electronics is not an exact science.Don't try to keep tolerances to narrow to start. Your going to find  every single component will have a fairly broad range of operation.
It's not magic,just physics.And its not hard to understand.
Most of all .Have fun with it. 
 

[techman-001]

Keep it simple, about+15/-15 and a separate +5 all rated about 1 amp will cover a lot of projects.  Search for 7815, 7915 and 7805 circuits.  Resist the temptation to overspec the PS to triple output 30A 3A.

Yes keep it simple because that way you will finish it in a reasonable time, but make it nice, and make it to last. Design a nice enclosure for it perhaps varnished wood (three-ply ?) with a brushed aluminum front panel and labeling, use the best terminals you can buy not some el-cheapo junk. Make the mains input a combo IEG plug with fuse on the rear. You may have problems buying a transformer with the windings you need, so use more than one, learn how to hand loom the cabling inside the enclosure. Learn what the expected dissipation is and make or buy the required heatsinks.

Spend some time making the 7815, 7915 and 7805 regulators easily replaceable, because you will blow one or more up at some point. Perhaps mounted on the rear of the enclosure ?

Make a etched brass or aluminum nameplate for the front, " Made by ..... in 2019" that way you'll know when you made it in another 40 years.

Make something you will be proud of every time you use it and in 20 years because test equipment is something you keep.

My *big* bench power supply weighs about 20 kg and has 0 - 30v @ 30A, 2 x 12v @ 1A fixed and one +5v @ 3A fixed. The heavy transformer has a single 0-30 volt @ 30A secondary winding which I regulate with SCR pre-regulation and linear second stages. The transformer and all semis are force fan cooled. I made it about 40 years ago and haven't used it in a decade. Nowadays a small mains inverter makes much more sense, but of course I didn't have access to that kind of technology back then.

[techman-001]

I think I'm going to build a power supply to start with and see where that takes me.

Oh, another one, since I joined this forum I see so many beginners pushed to start with a power supply project, seems we are really short of ideas here    For me a power supply is absolutely no fun, a top most boring stuff: can't play music nor make dance some remote stuff with it,  not even a way to have some fancy display, really nothing to play to the gallery (kids/wife etc.) 

A power supply is the foundation of electronics, no power, no electronics, so it makes sense to start at the beginning and work ones way up from there.

Assembling kits that look flash and impress people who don't know anything about electronics either, whilst good for stroking ones ego's isn't electronics, but it's probably art.

You may be looking for artblog.com ?

[icharters]

I think I'm going to build a power supply to start with and see where that takes me.

Oh, another one, since I joined this forum I see so many beginners pushed to start with a power supply project, seems we are really short of ideas here    For me a power supply is absolutely no fun, a top most boring stuff: can't play music nor make dance some remote stuff with it,  not even a way to have some fancy display, really nothing to play to the gallery (kids/wife etc.) 

A power supply is the foundation of electronics, no power, no electronics, so it makes sense to start at the beginning and work ones way up from there.

Assembling kits that look flash and impress people who don't know anything about electronics either, whilst good for stroking ones ego's isn't electronics, but it's probably art.

You may be looking for artblog.com ?

I'm inclined to agree.  Power is the basis for everything.  If I don't understand how to take power in, adapt it, make sure it's clean, distribute it, etc.  what hope do I have for doing anything else.  Also, I'm a low voltage technician by trade, and went to school for computer science (and continue to build stuff for fun), so just hooking up arduino development boards and adding peripherals and doing some light coding is not interesting as it isn't going to teach much if anything.  I want to design my own PCBs from start to finish and do cool custom hardware stuff.

I'm sure a power supply will keep me busy for a while.

[rstofer]

Jameco has a nice little +-15V power supply as a kit:

https://www.jameco.com/z/JE215-Dual-Output-Adjustable-Linear-Regulated-Power-Supply-Kit_20626.html

I built it for op amp circuits but I haven't used it yet.

Note that the output is adjtable from +-5V to +-15V.

Yes, it's a kit so there is no design involved but learning how it works is useful when it comes time to power other projects.  You can fall back on the kit design as a starting point.

Power is pretty fundamental yet somehow I did without a bench PS for about 60 years.  I used batteries, fixed supplies and, lately, wall warts.

[Siwastaja]

Getting off-topic, but I don't think building a "lab" power supply out of some 78xx regulators is necessarily a wise project. The reason is simple - it's not a lab power supply. It mimics the form, but not the function. For a psychological trick, if it makes you feel like you are in a lab, fine.

But for a lab supply to make any sense, it needs the basic lab supply features, meaning: adjustable voltage, and current limit. These are fundamental.

You can go surprisingly far without these features, for example, by using random wall warts, batteries etc. as explained by rstofer. Series resistors as current limiters. And so on. But then do you need the lab-supply-like neat box? I didn't, I played with wallwarts, batteries, etc. as well, for two decades I guess!

An actually proper lab supply, while useful even for a beginner, tends to be too difficult to build for one. The old magazines and the web is saturated with poor lab supply projects which just cause grieve, and for a beginner it's hard to know which one is bogus. Many beginners get thrown into compensation network analysis in their first project when their DIY lab supply is oscillating, while they could be building an internet-of-things automated coffeemaker out of hot melt glue, cable ties and duct tape.

I did do many LM317-based circuits whenever I needed adjustable voltages, but these were half-an-hour jobs. IMHO, it doens't make sense to spend a lot of time to refining casing and connectors for such a circuit. After all, these regulators ended up going inside the projects in the end. Wallwart, wired to a 7805 in a cardboard box, held together with hot melt glue. Great stuff.

But whatever floats your boat.

[techman-001]

Getting off-topic, but I don't think building a "lab" power supply out of some 78xx regulators is necessarily a wise project. The reason is simple - it's not a lab power supply. It mimics the form, but not the function. For a psychological trick, if it makes you feel like you are in a lab, fine.

But for a lab supply to make any sense, it needs the basic lab supply features, meaning: adjustable voltage, and current limit. These are fundamental.

You can go surprisingly far without these features, for example, by using random wall warts, batteries etc. as explained by rstofer. Series resistors as current limiters. And so on. But then do you need the lab-supply-like neat box? I didn't, I played with wallwarts, batteries, etc. as well, for two decades I guess!

An actually proper lab supply, while useful even for a beginner, tends to be too difficult to build for one. The old magazines and the web is saturated with poor lab supply projects which just cause grieve, and for a beginner it's hard to know which one is bogus. Many beginners get thrown into compensation network analysis in their first project when their DIY lab supply is oscillating, while they could be building an internet-of-things automated coffeemaker out of hot melt glue, cable ties and duct tape.

I did do many LM317-based circuits whenever I needed adjustable voltages, but these were half-an-hour jobs. IMHO, it doens't make sense to spend a lot of time to refining casing and connectors for such a circuit. After all, these regulators ended up going inside the projects in the end. Wallwart, wired to a 7805 in a cardboard box, held together with hot melt glue. Great stuff.

But whatever floats your boat.

Search this page, the only person who has said 'lab' power supply so far is .... you.

I said 'bench' power supply and recommended making a nice one as a first serious project, something to learn new skills in electronics, something to be proud of every time it's used.

Make it so nice and neat that you'd be proud to take it to show a potential employer your skills in enclosure making, layout, wiring and looming and general initiative.

You'd probably get that job over other candidates with their 7805 regulators glued with hot-melt to the inside of a used pizzabox.

[Siwastaja]

I don't know about replicating the same ages-old beginner project everyone else has done and proudly shown to their potential employers.  It's like having a kid, you are very proud of it (as you should be!), but everyone else has them as well and yours don't look that special to them.

Well, maybe if you are applying to a really "low-level" job, like in manufacturing? Otherwise, it doesn't show much passion nor expertise IMHO. Just an opinion, I'm entitled to one as well!

I mean, at the time you start moving from a hobby to profession (if that ever happens), the chances are you are already a few steps ahead of that power supply project, and could bring a more recent, more ambitious project that shows off your latest skills better.

The first project I brought with me to a job interview was a microcontroller controlled (before the Arduino era) 10-channel temperature sensor / regulator / heater control (for chemical process control) bare DIY etched PCB hotmelt glued to a plastic plate, and I showed off the UI, menus, temperature measurements... The interviewer wanted to measure his body temperature and it showed some 36.6 degC, success  . Made a better show than a 7805 power supply.

A guy I knew, an avid fresh hobbyist, made amazing progress in little time and I think he brought his IoT smart coffeemaker thingie (which automatically weighed the amount of leftover coffee in the pan, its age, reported it to the social media, and so on) in some job interviews IIRC. I'm 100% positive it's a better show about his skillset (ESP8266, strain gauge amplification, AVR, ADC, internet protocols, product design and prototyping), than being able to neatly route wires to a 7805 and buy expensive-looking connectors from a distributor. This was something he built in less than a year after starting from scratch!

Note, no offense intended. Just wondering your assertiveness on this. Clearly it's an important project for you, and I appreciate that. Building a bench (sorry) power supply is not a bad beginner project. But it's overrated, IMHO, and the functional end result tends to be disappointing. The remaining effect is psychological, and this means it's very much personal with no ground truth. Whether others (such as employers) find the power supply project interesting, is up to a debate. Having done a few interviews, I would prefer the bench supply builder over someone who's done nothing, but would prefer a hotmelt glue IoT coffeemaker crazy scientist over the power supply builder if the job requires innovation and design experience on modern fields. For a production work position involving tying cables, I would pick the one who showed a neat power supply, any day - he/she is probably a very reliable and dependable person.

[tggzzz]

Thanks for the input, guys.  I think I will take the advice here and learn things on an as needed basis, and take a more scientific approach to learning it when i really want to deeply understand something I'm working on.
 
I think I'm going to build a power supply to start with and see where that takes me.

Good for you, and ignore the naysayers.

Do what you want to do, but be realistic about how much it will impress an employer. For example, if a potential employer is interested in real-time software, for example, it will be of less interest.

If you demonstrate that you:

• thought about your objectives and non-objectives, and the reasons for setting them
• thought about the corner cases of what could go wrong; many people think how things work, few about how things fail
• implemented and tested it
• found where it didn't work well enough, and improved it
• can describe why it worked well enough
• can describe mistakes you wouldn't repeat next time
• can describe how you might extend it, if the need was there

then most employers will be more impressed.

You could also assess an employer, based on whether they ask you interesting relevant questions about your project. If they don't ask interesting questions, work out why. If they only ask their own questions (e.g. what is an ACID transaction) or trick questions (why are manhole covers round), then assess them accordingly.

[Siwastaja]

ignore the naysayers. Do what you want to do, but be realistic

Cut like this, this is the best generic advice imaginable.

[techman-001]

I don't know about replicating the same ages-old beginner project everyone else has done and proudly shown to their potential employers.  It's like having a kid, you are very proud of it (as you should be!), but everyone else has them as well and yours don't look that special to them.

Well, maybe if you are applying to a really "low-level" job, like in manufacturing? Otherwise, it doesn't show much passion nor expertise IMHO. Just an opinion, I'm entitled to one as well!

I mean, at the time you start moving from a hobby to profession (if that ever happens), the chances are you are already a few steps ahead of that power supply project, and could bring a more recent, more ambitious project that shows off your latest skills better.

The first project I brought with me to a job interview was a microcontroller controlled (before the Arduino era) 10-channel temperature sensor / regulator / heater control (for chemical process control) bare DIY etched PCB hotmelt glued to a plastic plate, and I showed off the UI, menus, temperature measurements... The interviewer wanted to measure his body temperature and it showed some 36.6 degC, success  . Made a better show than a 7805 power supply.

A guy I knew, an avid fresh hobbyist, made amazing progress in little time and I think he brought his IoT smart coffeemaker thingie (which automatically weighed the amount of leftover coffee in the pan, its age, reported it to the social media, and so on) in some job interviews IIRC. I'm 100% positive it's a better show about his skillset (ESP8266, strain gauge amplification, AVR, ADC, internet protocols, product design and prototyping), than being able to neatly route wires to a 7805 and buy expensive-looking connectors from a distributor. This was something he built in less than a year after starting from scratch!

Note, no offense intended. Just wondering your assertiveness on this. Clearly it's an important project for you, and I appreciate that. Building a bench (sorry) power supply is not a bad beginner project. But it's overrated, IMHO, and the functional end result tends to be disappointing. The remaining effect is psychological, and this means it's very much personal with no ground truth. Whether others (such as employers) find the power supply project interesting, is up to a debate. Having done a few interviews, I would prefer the bench supply builder over someone who's done nothing, but would prefer a hotmelt glue IoT coffeemaker crazy scientist over the power supply builder if the job requires innovation and design experience on modern fields. For a production work position involving tying cables, I would pick the one who showed a neat power supply, any day - he/she is probably a very reliable and dependable person.

Mr  Siwastaja, let me say here and now, it's a pleasure to debate with you, and I also mean no offense to you. I have observed that you're a very knowledgeable person from reading your many posts here.

We probably don't disagree very much imho, but different paths lead to different life experiences hence our discussions so far.

I'm sure that a solidly designed bench psu would be a lot more useful to most electronics beginners than a IoT smart coffeemaker, but opinions may differ. In any event the bench PSU is just the *first* project, and on that foundation many other things would be built and benefit from previous skill learned.

Great devices don't just pop into existence from nowhere, they are hard fought for, require lots of skill, training, hardwork and many redesigns until everyone hates the sight of them.

I'm not hung up on PSU's, the OP asked what parts/resources he needed to get into electronic design and I suggested the PSU as a good place to start because then what he needs will become clearer.

Now do you really think a interviewer will be interested in your friends IoT smart coffeemaker thingie built with a ESPxxx ? He's probably the 3rd one he's interviewed with the same thing that week as IoT is all the rage with the kids these days. You claim the PSU builder with his gleaming, varnished home made psu that demonstrates a half dozen mastered relevant skills would fail to impress ? I bet hes the *only* one with those skills the interviewer has seen.

Of course we both know that interviewers are usually clueless and if they don't call in a engineer for a look at whatever you have brought in, you've wasted your time anyway, especially if the engineer thinks the ESPxx is the last thing he would ever design with or that it would fail to be approved for use in R&D or they're already having great success with LoRa on CortexM.

On the other hand, the guy with the PSU could easily find himself as a tech in R&D for that company, making up gear for the chief engineer just like I did 40 years ago. The gear was a prototype nucleonic iron ore flow detector using a National PACE 16 bit development system. The engineer would give me the circuits and say 'make!", the rest was up to me.

Finally, if anyone around here is building gear on cardboard with hotmelt, it will be you and me, because those skills require 40+ years of R&D experience learning what we can get away with and what we can't, it's not for beginners!