Electrical Engineering

« first day (3641 days earlier) ← previous day next day → last day (1299 days later) »

1:21 AM

Here's my takeaway from our exchange about C++ for microcontrollers.

The majority of C++ programmers aren't microcontroller programmers. They lack knowledge about how to write efficient code for small platforms. As efficient as plain C or close to that.

Many EE types don't grok C++. EE types, however, frequently have to write code for microcontrollers. So they stick to plain C. **

There isn't anything fundamentally wrong with C++ for microcontrollers (compared to C) provided that:
(1) The firmware authors know C++. They also know what C++ language features to stay away from on microcontrollers. The number of these features is manageable: somewhere between 5 and 8.
(2) The C++ compiler is good.

These conditions rarely come together in real world microcontroller projects.
As a result, C++ on microcontrollers remains... a novelty item.

1 hour later…

Nick Alexeev

2:27 AM

** C++ is an acquired taste. Likewise, the majority of C++ programmers don't grok EE and physics.

6 hours later…

Lundin

8:35 AM

@NickAlexeev Yep, that about sums it up. Overall in my experience, it is much easier to take an EE with a bit of assembler knowledge and turn them into a competent embedded C or C+ programmer, than to do the same with a PC programmer who got tons of high level knowledge. The key is to have them actually understand (roughly) what actual machine code each line of C or C++ results in.

Like understanding that adding 2 double together on an Arduino results in a massive software floating point library getting linked and inlined with the code, slaughtering performance, RAM and flash all at once. Because it doesn't have a FPU and it's a lousy 8 bitter that hates large numbers in general. Because... it is not a PC.

5 hours later…

W5VO

1:20 PM

Heh, l love the header... "if we were not beggars but choosers...."

3 hours later…

TeilZeitGott

4:47 PM

help help beggers and choosers:

Q: Finding EMF from the given graphs

Here are the graphs of flux vs angle and EMF vs angle. From it, I have to derive the following relation (expected answer) between EMF -Ui and change in flux with respect to time. My approach: I find the function for every piecewise continuous graph which gives me the following result: Φ=-B∙A, ...