#Went from not knowing much about electronics to now the fourth revision of this PCB in less than 6mo

42 messages · Page 1 of 1 (latest)

echo dove
#

Working on a very large, "whole life" project that needed some special electronics, so instead of paying just a little bit of money to have something designed for me, I instead spent a lot of money to learn the skill myself. Money well spent. Just finished this today, and while it's not 100% tested (the ethernet subcircuits and OLED still need to be tested after some modifications) the board is nearing completion. Just wanted to share as I personally think these designs look really pretty, even if you don't know what it's all doing 😄

image.png
#

Some 3D views:

image.png image.png
#

It's also open source.

frank wadi
#

But well done for learning a new skill!

craggy crane
#

Very impressive work.

echo dove
#

Thanks! :)) Will post pictures if it works and when it gets assembled

#

otherwise there will be a rev5 😂

craggy crane
#

How was your experience with KiCad as a beginner?

jaunty cedar
#

This is awesome. Mind sharing your resources for learning electronic design? I only know a bit about the basics of it and it's something I have missed knowing better many times

echo dove
echo dove
# jaunty cedar This is awesome. Mind sharing your resources for learning electronic design? I o...

Honestly? ChatGPT helped a lot at the beginning. I didn't trust any of the answers it gave me but I would ask it questions and then use the responses as keywords/search terms. Helped a lot. Also, understanding what you need, and then going to e.g. mouser.de and finding things that seem right, reading the datasheets, etc. Reading datasheets taught me most important things, and most of them for e.g. the ICs/MCUs tell you what kind of circuits you need along with the IC, so you can then look up why you need them, etc.

#

There's also an EE discord that I go to https://discord.gg/mmEamQFZ8K - I ask a bunch of questions there and they're very helpful. Strong warning though, they're a bit, uh... problematic in the off-topic channels.

#

Also, my first boards were terrible. Start small, get the process down (component selection, schematic, PCB layout, order the boards/parts, assembly, test) with something easy and simple first, just so you get an idea of how it all works.

#

It's a LOT of referencing, a lot of looking up, etc. Knowing the keyboard shortcuts in Kicad helps (e.g. d while mousing over a component in the schematic editor to quickly open the datasheet) saves you a lot of time and effort.

#

Lots of trial and error, having good tooling IRL (a good multimeter and a good soldering iron) plus learning the quick tricks for doing a lot of SMD soldering quickly (i.e. a 25$ skillet and some aquarium sand off of amazon + a cheap tube of T3 soldering paste is what I use and it works well. Doesn't come out super clean but gets the job done), some needle-sharp tweezers, and depending on if you're paying attention to the component sizes you pick out (some are, literally, smaller than dust) you'll want a microscope at some point. Really helps, especially with eye strain.

jaunty cedar
#

Awesome, I'll look into it when I have the time and a specific project for this, at least I have the multimeter and the iron 😅

echo dove
#

Ping me if you ever have questions 🙂 I'm not an expert but can probably point you in the right direction!

#

I'm focused pretty much entirely on digital, not analog/RF - with a few exceptions (that aren't really exceptions, but exception enough to warrant stating they're exceptions). So I can help if that's the case.

#

Anything radio-related is out of my depth though.

#

Also, much more difficult to design.

jaunty cedar
#

Yeah, I know analogue electronics in general is a pain, which sucks because for the projects I would have needed it, it was very much around biosignals (so deep into AE)

echo dove
#

Had to make a programmer board. The pieces will stack up using M3 spacers and connect to the programming pad via test probes. The extra ... strange pieces attached to it will break apart and form the intermedia layers, bracing the probes into the programmer (hopefully, if I got my CAD right).

Hopefully JLC will approve the manufacture, this is probably going to be a hellish thing for their CNC to do.

image.png
echo dove
#

There's no way they're going to accept it 😂 they didn't even import it correctly

echo dove
#

So they're producing it

wraith ocean
#

tbh I kinda figured they would...weird shaped PCBs like that are actually not all that more difficult to produce if you're CNCing it anyway

#

you clearly designed it with the tool diameter in mind so there's no reason for them to reject it

#

whoever runs these is going to be very confused about what they're for though lmao

echo dove
#

next revision of the main board

WhatsApp_Image_2023-06-02_at_18.33.53.jpg
#

much less chaos happening than before

rough bronze
rough bronze
#

It is also good idea to pick a micro-controller architecture that you're comfortable with, build a breakout board for it (power, clock signal, I/O) and write/test a standard library for that architecture before using it in an application. This way you don't have to rely on someone else's code and you now have a platform (code, reference designs, partlist etc.) which is reusable in the future. Make sure the parts that aren't easily exchanged (multi-pin integrated circuits with unique functionality) aren't end of life, critical to some large manufacturers production runs or have similar availability constraints.

#

Assuming you've internalized the basic theory, The Art of Electronics 3rd Edition is a good reference. Physics of Semiconductor Devices is also very beneficial and The Car Hacker's Handbook if you're going to dive into things that move.

#

I've got PDF copies of all three hosted (on a super slow site) if anyone wants them and can't find them via search engine.

#

Too big for Discord.

#

Out of interest, what is the purpose of Oro?

#

Memory safe embedded OS for stuff and things?