PCB milling and circuit projects

  • Guest, it's time once again for the hotly contested and exciting FoH Asshat Tournament!



    Go here and fill out your bracket!
    Who's been the biggest Asshat in the last year? Once again, only you can decide!

Erronius

Macho Ma'am
<Gold Donor>
16,456
42,340
I was talking to people at my school about etching PCBs (we still have a lot of supplies in a storage room that have been there forever) but people were telling me that we could just use some of the small mills that we have to do that, instead of fudging around with chemicals. I've been looking at Eagle PCB and rapid-pcb.com, but I haven't taken the leap into any of that yet. Does anyone have any experience with anything g-code at all or STL cad files for something like this? Our CAD and prototyping people said they'd be more than happy to help, but I've zero experience in this and I think people will be looking at me to come up with circuit ideas and PCB layouts for them to do the actual machining from.

Also I'm looking for projects to make that go a bit beyond a breadboard and have some utility or at least could stand as part of a centerpiece type project, if anyone has any ideas. I was looking at things like nixie clocks, tube amps and analog synths but I feel like I'm veering off into an odd direction with those. The other direction I felt pulled towards was rasberry pi/arduino projects or other things microcontroller~ish but I'm afraid that might turn out to be a really narrow focus as well.

Finally, does anyone have a suggestion on non-curriculum circuit collections along the lines ofEncyclopedia of Electronic Circuits Volume 1? I knew someone who had a copy and had a chance to thumb through it, but after reading some of the reviews online (especially the later volumes) I'm getting the impression that there's a lot of mistakes, there are some circuits that don't work at all and that some people don't seem to find the books to be reliable/accurate. I just wanted to have some sort of reference of useful/interesting circuits without having to sort through the entirety of the internet.
 

gogusrl

Molten Core Raider
1,359
102
How about making some keyboard pcb's ? Custom made keyboards are all the rage now and there are a lot of options out there with practical usage (at least for me and a few others). I'd pay some serious +nets for some ergodox or phantom pcbs.

For example :

http://ergodox.org/

Projects - GeekHackWiki
 

Noodleface

A Mod Real Quick
37,961
14,508
What's your hardware/software background and what are you trying to accomplish? Just saying you want something to be the centerpiece of a project is one thing, actually narrowing the project down will help you better decide which direction to go.

I am a RaspberryPi hobbyist, so I can answer anything about that. Know a bit less about Arduino, but it's close.
 

Erronius

Macho Ma'am
<Gold Donor>
16,456
42,340
How about making some keyboard pcb's ? Custom made keyboards are all the rage now and there are a lot of options out there with practical usage (at least for me and a few others). I'd pay some serious +nets for some ergodox or phantom pcbs.

For example :

http://ergodox.org/

Projects - GeekHackWiki
I honestly hadn't ever seen anything about custom keyboards, but then again I haven't followed anything really hardware related in years either. Do people doing this have to create their own PCBs from scratch or do people pass around existing PCB designs?

What's your hardware/software background and what are you trying to accomplish? Just saying you want something to be the centerpiece of a project is one thing, actually narrowing the project down will help you better decide which direction to go.

I am a RaspberryPi hobbyist, so I can answer anything about that. Know a bit less about Arduino, but it's close.
I'm mostly looking for things to branch out from my electronics courses with. I'd brought up Arduino and Pi to one instructor and he suggested I wait until this spring semester for a Microcontrollers course. He never seemed very impressed with Arduino and said he preferred Rasberry Pi, but he also was kinda /shrug about both and said we could just make our own versions. It's hard for me to say how practical an idea that is though because he has like 40 years experience and used to prototype quite a bit, so it might be something more along the lines of him been quite able to do something like that but at the same time not be remotely practical for us. I know we're going to use some flavor of atmega this spring but I don't remember what exactly.

I haven't taken programming since VBasic a long time ago, and I doubt PLC ladder-logic really counts. I'll probably be taking a C language this spring and we'll be doing assembly in the Micro class (along with CAD, welding and some capstones because I keep running into projects where I wish I had various supporting skills).

Part of me wanting to branch out and find projects is due to me wanting to push forward beyond some of these classes, especially since I might be transferring into programs at other schools. I'm a little afraid that if I don't work on stuff on my own that I'll run into issues later (forgetting stuff or being behind on a practical level). Another part of it is this school has a lot of available tech but there isn't always a lot of inter-department cooperation on projects. So like with the PCB deal...we have old chemicals and kits, but no one I know has touched the stuff in years, people in my department know in their own minds that milling PCBs can be done but AFAIK no one has ever approached the CAD/prototyping guys about it before now. For us in an 'industrial' department there probably isn't much need for custom PCBs, but part of me feels like we have the potential to do some really cool stuff and any experience/knowledge I pick up along the way can't possibly hurt. Our CAD guys actually sound pretty interested/excited but if I don't take the lead then I don't think anyone else will.

The "centerpiece" thing is kinda like...I can build stuff on perf or breadboards, power supplies or amps or whatever else and honestly I'm not sure that anyone else outside of a particular class would ever give something like that a second glance. But if it's a circuit that's built into a finished product, like for a keyboard like Gogusrl mentioned or nixie clocks or a robot arm...that's something that people can look at and be interested in as opposed to"oh hey, a breadboard with a metric fuckton of stuff on it". So the departments at school would have stuff to show off, and maybe I could make gifts for friends or stuff I could use myself.

Anyways I need to crash again, that's enough rambling for now
 

Noodleface

A Mod Real Quick
37,961
14,508
For the RPi and somewhat for the arduino I would hold off until you take C. You won't use assembly on them unless you really want to, although asm gives you a much better idea of how things work at a much lower level (moving bits, jumps, etc).

The RPi will essentially function as a small linux PC that you can also attach things to (the GPIOs). I've used a few things like cameras, speakers, GPS modules, and a few other things by wiring up and programming their uses through C - so it will be good to take that course before jumping in.

The RPi is way more powerful than any Arduinos right now, the downside is it's not as 'easy' to program as Arduinos. Arduinos have a lot of very useful libraries that the RPi just doesn't have yet. Also, the Arduino has a lot of proprietary devices you can connect very easily like controllers, GPS, ethernet shields, etc.

I'm a big fan of the RPi, but I also write in C everyday at work so it's a bit more natural for me.

As far as etching PCBs you probably won't ever really see people doing that on their own. Even at school we farmed it out to external companies.
 

Erronius

Macho Ma'am
<Gold Donor>
16,456
42,340
As far as etching PCBs you probably won't ever really see people doing that on their own. Even at school we farmed it out to external companies.
After taking a capstone course over the summer and trying to prototype and etch+mill some boards, I think I came to the same conclusion. It was a learning experience. Our CADD folks had no idea WTF a gerber file was (EagleCad), while I had no knowledge of .STLs and g-code. First I took a completely unnecessary detour with trying to convert filetypes towards .STLs from gerber files. Then I got to experience the joy of incorrectly trying to output a file (I'd have to look it up specifically but kept having java applet crashes /boggle) and only having a small portion of the board rendered into the code itself. CopperCad can eat a dick BTW. Then we finally had success with putting PCB_Gcode into the ULP and just $*%#&^# exporting the damn gcode without needing to go through any bizarre string of conversions.

Of course I knew none of this, so mistakes were made.

We got a board milled, but A) since we drew the single-sided board on the bottom layer it automatically mirrored it, but B) we didn't think of that so we we clicked "mirror" and ended up flipping the bottom layer such that the IC would have been in backwards, or we would have had to mount it from the copper side C) we had to use a mill with a relatively low RPM of 3k-5k, which meant it took like 4-5 hours just to mill the traces D) good luck getting your board level.

In an 8-week summer semester we had so much time taken up by slogging through trying to get the milling straight that I still haven't had time to try etching, though we have the mats. My capstone report essentially became"Why you should just make sure that your files and layers from EagleCad are correct and send that shit to someone to make for you".


Oh, and I took that Microcontrollers class in the spring semester, and I can't say that I am a big fan of assembly. Interesting to play with and I'm sure that it's "light" so to speak, but FFS. Timing loops and NOPs, /sigh