Once I finally decided "ah, what the heck, I'll take a shot at the Hackaday Trinket contest," my actual project idea came pretty quickly. Here's the in-depth explanation of what I'm doing and why I'm doing it!

The problem:

I spend quite a bit of my "hacking" time working on high altitude balloons with my school's aerospace engineering department. It's an awesome project, amazingly fun, we get to build actual flight hardware, all the bells and whistles. Of course, not everything always goes to plan (when does it ever?). One of the larger issues we end up combating is faulty or incorrectly set up sensors. The lab we use as our mini-makerspace has been home to the HAB program since its inception more than a decade ago - and several of the leaders of the program in that time were notorious hoarders. The end result? Literally dozens of sensors, mostly simple analog breakout boards, that may or may not still work. Since no one wants to spend money on a new sensor, these things get used and abused way too many times to count. I'm fairly certain that I've even seen a payload that consisted entirely of discontinued products. I wish that were hyperbole.

This is all well and good except for a few inconvenient tidbits. First, the people who usually grab old sensors are by and large our newest, most inexperienced groups. They're usually cobbling a payload together with what bits they can find in the lab and what code they can find online. Almost universally, this means that they don't really bother testing or calibrating anything. The most common excuse, which persists even after being shown how to test or calibrate a sensor, is "That takes too much time, I don't want to set it up!".

My solution:

My solution to this problem is simple: create a plug-and-play sensor tester. For the inexperienced new guys, the setup time, and especially the super scary code (the horror!) will be taken out of the equation, and they'll have a better shot at a successful first payload. And trust me - a successful first payload is key to retaining interest! For the program's veterans, including me, we'll have a quick setup to test the piles of sensors lying around and perhaps thin out the collection a bit. Additionally, even we complain when we have to breadboard a circuit just to test something for 30 seconds. Cutting setup time increases test time, and more testing is always good.

Another advantage is calibration. No two sensors are perfectly alike, and it can show! One of the most striking examples of this that I've seen happened just last month. An analog accelerometer, using the datasheet's conversion from voltage to G force, reported Earth gravity as 1.2 G. That's...not exactly insignificant, especially when the sensor is being fitted to a prototype shock absorber! However, once we knew about that inconsistency, we could easily correct for it in our code. The downside was that in order to calibrate, we needed to assemble the payload, run the Arduino for a minute or so, disassemble the payload, remove our logger's SD card, and pull the data off of that. Wouldn't it be simpler to just plug in to a pre-built tester and get instant results? I certainly think so!

As you can see above, I've already specced out pretty much everything for this project. Tomorrow, I'll write up another update with an in-depth look at the design so far.