Comment by mlsu

Comment by mlsu 6 days ago

6 replies

For this, what takes a while is to just tinker and fry components several times. Get a breadboard, get several sensors, try to design something and iterate on your design. Plan to fry sensors and IC's. Also helps to read some basic electrical theory and know what the role of different components are.

The way I got proficient is with hobbyist PCB design. What helped me is starting with schematics and datasheets and planning to finish with an assembled board. I started designing PCB's and having them assembled with JLCPCB (quite cheap: $20 or so for a run of 5 boards; $120-$150 fully assembled). I fried 2 boards before the 3rd rev booted up, then from there it's optimization. I consider the $200/mo or so in PCBA, whether boards work or not, to be my "EE education" -- cost efficient compared to university fees! And $200 is sort of like the "exam," it's costly enough to make me really think twice about component selection/placement/etc.

Not saying that's the approach you want to take because that might be hardcore / not someplace you want to get to. But I spent a long long time really wondering how electricity really works and like why you need capacitors, inductors, op-amps, etc. It never made sense to me until I created my own schematic, chose my own parts, and understood why I chose the parts I did and connected them the way I did.

iLoveOncall 6 days ago

This just reinforces the fact that it's inaccessible. There's no way I'm literally throwing $200 a month in the trash on a hobby.

  • yetihehe 6 days ago

    $200 is for assembled boards. I learned electronics and spent about $200 on it in two years, that includes cheapest soldering iron. Don't order assembled boards when you are starting. Order cheapest bluepill (STM32F103C8T6) or non-original arduino clone and start on breadboard with that. Make pcb's only when you're ready to learn more. Expect that your first one will not work or will require some "rewiring", but second one may already work. You might start with some cheap through-hole components, they are a little easier to re-wire or re-solder, it's a good idea to put your first microcontroller in socket.

    • iLoveOncall 6 days ago

      Ah alright that makes more sense, I missed that part, my bad.

  • jocaal 5 days ago

    I'm a EE masters student and I also want to reinforce how inaccessible EE is and that I don't really recommend it as a hobby. EE is a very mature field and it's very math heavy for a reason. The second you move past the hobby boards, stuff becomes really difficult and really expensive really fast. If your end goal is to create toys for kids, then it's fine as a hobby. But without the formal training and lab access you are going to struggle to get past that point so it's pretty much impossible to turn the hobby into something more. Unlike software, where tinkering genuinely has the possibility of turning your side projects into careers. Hell, if you don't live in EE hotspot locations, I wouldn't even recommend it as a career anymore. Software is where it's at, even in the age of AI.

  • eternityforest 5 days ago

    That's only for the heavy analog stuff. If you're into the more modern digital stuff, you basically never for any reason need to breadboard prototype, everything can be done with I2C modules and the like.

    Burning a part is incredibly rare with this kind of stuff, if you're willing to put in the time to learn about it before actually building it.

blankx32 6 days ago

did you understand theory deeply first like Kirchhoff , node analysis