You learn at least Altium, PADS, Eagle and some open alternatives. You learn the specific quirks of all those software. You hate every hour you have to fight with PADS' decades-old bugs and "features".

In addition to LT Spice you probably learn a more specialized simulation software. Who knows, maybe you have to design power elecronics. Maybe antennas. Maybe both. Maybe in the same system. Maybe you are specialized in power electronics and you have to integrate another specialist's design into yours.

You learn about all the basic components, basic circuits and how to apply them. And yes, to "read some data sheets". You learn about the most common IC's and how to apply them, and when not to. And so on. You learn about EMI. You learn about trace lengths, placement, and their myriad of requirements in a multitude of scenarios. You learn about multilayered PCB design and its requirements. And so on.

You learn about having to minimize cost by reducing the components to their absolute minimum. You learn about having to minimize space due to real life demands of where your PCB has to fit. Maybe you even specialize in PCB design, and you apply a specialist circuit designer's schematic to the real world. Or maybe you are said specialist designer. In any case, at some point you need to know both and neither is easy.

Inevitably when you have to lock in your design, the prototypes are manufactured. Perhaps you need to assemble them yourself, certainly at least partially. You may even have just a partial device at first, which you still need to be able to wake up. Without fail, the real world reveals its ugly face to you. Despite all your simulations and meticulous design, something doesn't work like it should. You likely need to have mastery of a table full of expensive measurement equipment, and you need to know what to look for and how to look for it. After you diagnose the issue, you need to know how to fix it. You probably have to fix it manually, likely involving precision soldering work to sub-millimeter pads. After the fix, you don't hit "recompile", you fix the schematic, then the PCB, then wait for weeks or months for the next prototype round. Until then, you need to continue testing the device with hand-fixed prototypes. You need to instruct other people on how to do the same modifications.

Who knows, maybe a pandemic happens, maybe also a war that affects global economy. A chip you used becomes unavailable. There is no drop-in replacement. You have to at least replace the chip, and perhaps the circuit surrounding it. You don't just hit "recompile"...

All the while you likely have one or multiple microcontrollers in the device. You need to interface with the SW architects as well as the developer side. You need to be the datasheet for the software developer - likely you even need to write the datasheet.

All that jazz in addition to the regular stuff everyone has to deal with in one form or another.

I'm not an EE, I'm an Embedded Systems and Electronics Engineer. So I don't know 100% how to explain what they do, although I have some peripheral knowledge. Which is exactly I wrote this. I have just been in my safe firmware developer's bubble, trying to support the people designing the hardware so that they can get their side in order with the MCU's doing what they should. I have immense respect for the people who are able to make the physical side of what I write the code for. Those circuits certainly don't just "pop out".

Both sides always have something to be mastered. Both sides have their different flavors, too, but neither one is easy. Both sides have to learn and apply. Funnily enough, both sides have clear analogues to eachother, and very similar problems, just in a different plane of existence. Oh and this is just one facet of EE like others have mentioned.

I could never be a hard EE.

user:	atune
created:	March 27, 2017
karma:	24