3 ms·

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 a

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.

lmilcin 6y ago
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.
p0llard 6y ago
> 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.
- PascLeRasc 6y ago
  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.

lmilcin 6y ago

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.

p0llard 6y ago

> 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.

PascLeRasc 6y ago
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.