Engineers should understand the underlying physics plus know how to use it for something useful. There are degrees such as "electrical engineering technology", which cut out the harder fundamentals (and advanced classes that require it) and just focus on things you can do with current technology. Such skills become obsolete faster as technology changes.

I'm not sure I completely agree with this: electronics refers specifically to (essentially) anything concerning the control/emission of electrons (so transistors, thermionic valves, etc.); it's not just "advanced electrical engineering".

The divide you're talking about exists in both electrical and electronic engineering, if they are to be considered separate disciplines: in electronic engineering you have both the solid state physics required to understand semiconductor devices, and the layers of abstraction used to design analogue circuitry; in electrical engineering you have all the theory of electromagnetism, and the layers of abstraction used to design electrical machines.

The divide you're referring to is real, but it definitely isn't the electrical/electronic divide.

Electronics is about building useful circuits. Do you really need to understand what precisely happens within any of the parts? No, not really. I can't imagine what happens to electrons in MOSFET nor do I care. For my purpose it represents some transfer function which is what I use to build the circuit. When I think about OpAmp, the physics is not one of things I am thinking. My mental model of OpAmp has no physics in it and it is about relation between input and output signals.

Think this way: do you need to understand how a complex IC part works? No, you don't. You read the manual and learn that if you put something on particular inputs you will get something on outputs. You leave designing the internals to others. You take parts other people built and solder them to long pieces to copper glued to PCBs.

That's really most of electronics.

You need to know a little bit of physics. You need to appreciate some phenomena like losses, noises, you need to know what happens at high frequencies, maybe you want to understand how heat is conducted away from your parts, etc. You need to know couple extremely simple formulae and laws (laughably simple to any physicist interested in the matter). Other than that your parts function as tiny little lego bricks that are transfer functions to affect how your circuit works.

> Electronics is about building useful circuits.

This is just factually inaccurate; electronics is the field which relates to control/emission of electrons.

If you build a "useful circuit" which does not use any active components, you aren't doing electronics.

If you're applying solid state physics to model transistor behaviour, you're doing electronics.

The terminology has absolutely nothing to do with levels of abstraction.

I understand what you're saying, and agree that there is a broad range of levels of abstraction within electronics, but the division in abstraction is unrelated to the division between electrical and electronic engineering.

There are multiple definitions of electronics. If you are physicist you might envision electronics as control of emission of electrons. If you are an actual engineer you will have very different definition and understanding of what electronics is.

It is like calling computer scientist experts in software development. No, knowledge of algorithms is far, far from software development which is also about human interaction, project organization, and many, many other things.

Go google "electronics definition":

>noun

>the branch of physics _and technology_ concerned with the design of circuits using transistors and microchips, and with the behaviour and movement of electrons in a semiconductor, conductor, vacuum, or gas.

So no, it is not obviously just about control/emission of electrons.

> the branch of physics _and technology_ concerned with the design of circuits using transistors and microchips, and with the behaviour and movement of electrons in a semiconductor, conductor, vacuum, or gas.

Yep; it is explicitly is talking about either active devices or literally cases where the topic of focus is control/behaviour of electrons.

> If you are physicist you might envision electronics as control of emission of electrons. If you are an actual engineer you will have very different definition and understanding of what electronics is.

It has nothing to do with the level of abstraction at which you're working, which is my entire point; whilst an engineer may very well not be considering the behaviour of electrons in day to day work, if she is doing electronics, she is working with devices/technologies which involve the control/emission of electrons. If she is working on, for example, an induction machine, she is not doing electronics.

Again, I'm not contesting that people work at different levels of abstraction within a field, just pointing out that the term "electronics" as opposed to "study of electricity" has absolutely nothing to do with abstraction; it refers specifically to whether <active devices/transistors/semiconductors/control and behaviour of electrons/whatever you want to call it> is involved.

EE degree here to back you up. I believe you're correct, electronics is to electricity what software engineering is to CS. It's the practical application with as little depth as you can get away with. For example there's some bits of RF you'll learn when you look at 45 vs 90° PCB traces, but you don't need finite-difference time domain equations to understand what's going on.

Exactly. I know enough physics and mathematics to know that what is used to create designs is laughably simplistic and oftentimes incorrect knowledge. But it doesn't matter, we are not pushing physics with the designs. We just want enough heuristics to be able to build something that works.

Electronics engineers don't go wielding equations to build circuits (well, except maybe Ohm's Law) but rather rely on intuitive models they have in their brains of what components do when put in a specific place in the circuits.

This is necessary shortcut because otherwise even simple circuit can become extremely difficult to understand if you start from first principles.

> electronics is to electricity what software engineering is to CS

I would respectfully disagree with this assertion.

I'm pretty sure my colleagues who are designing novel FinFET transistors for low noise applications are "doing electronics", and I'm also pretty certain that they're considering the underlying semi-conductor physics in some pretty serious depth; they're definitely not trying to work with "as little depth as [they] can get away with".

Electronics as a discipline encompasses people working at many levels of abstraction, including those working at a very low level. I think transistor designers would be very amused (or perhaps offended?) if you were to claim that they weren't doing "proper electronics" because they're actually thinking about things in depth in a very analytic way.

Oh I wouldn't ever claim component design is anything but complex. I was just trying to make a distinction between high-level fields and something like embedded hardware integration where you can treat each component as a "black box" of sorts and just care about I/O and operating characteristics. I don't mean to diminish your colleagues' work, it sounds fascinating.