I’ve made a few PCBs for myself, but I needed a new challenge. Could I make something for other people, I wonder? Well, it would have to actually work properly, for a start. And it would have to be safe (depending on what it’s doing). A quick call out on my regular IRC channel provided a possible project: a little controller board for someone’s around-the-monitor Ambilight setup.
He’d had some jerry-rigged setup with twisted wires and dodgy solder joints, so I offered to make a proper PCB for it. He was going to make a box for it with a 3D printer as well to make it nice and tidy. The LED strips were standard WS2812b strips of the kind you can buy by the metre.
There were a number of things to consider with this that made it a bit more complex than my earlier projects. Firstly the mechanical integration, requiring some placement of screw holes and orientating components for appropriate hole placement and cable access.
Secondly, and more importantly, there was the current and heating requirements to think about. At full brightness my friend estimated that the entire strip around the monitor would pull around 6 Amps. This meant that the PCB would possibly run quite hot, and it also meant that the cables and connectors coming off it needed to also be able to support this too. Pin headers, apparently, may not be right choice here.
The schematic was fairly simple, as you can see:
And it was a fairly simple job to lay out the components on the board, allowing for a cable to come out of either side for the strips, plus power and a space for the Arduino Nano that would run the whole thing to sit.
But there was still the issue of the current draw. How wide should I make my tracks? I tried to join some weird polygon shapes together around the pads and had all sorts of worried about clearance and angles and pinch points that could get hotter then the surrounding areas, all of that stuff. And then I thought I better ask for help in case I kill my friend, so I went away to the Something Awful electronics thread for some guidance. Turns out that after using entire ground plane fills on the back side of my boards for so long this was just a similar issue. I ended up making the entire top plane a +5V fill: Problem solved.
Here’s the finished board, without any components in this time (points for spotting the reference):
The enclosure my friend designed is here for completeness as well, and I’m happy to report it all worked fine with the heat dissipated nicely. In the end he stayed with the pin headers and used some appropriately thick cables and everything ran fine - not too toasty.
I’m now always bugging friends for new ideas. Still addicted to this a bit!
Skip to main content
Comments
Post a Comment