Hardware Lyfe

Hi – we’ve been a little distracted by the holidays and more recently, the terribly cold weather (for Texas and me—some of you are made of sterner stuff). In spite of my preparations, we didn’t have hot water for two days, but it’s thawed now with no damage. This is when I break out the Twines to monitor areas of concern.

screenshot of Twine dashboard


series of printed case prototypes

I’m asymptotically approaching final mechanical design. This is the part of the process that separates us hardware sickos from the rest. Iteration cycles bring progress so slow that it can be disheartening. And we’re on the verge of committing tens of thousands of dollars for injection molds we can’t significantly change. I’m experienced with this but that doesn’t mean I’m fearless.

Part of the recent slog is conversations with our injection molder to optimize the design to bring down the tooling cost and improve moldability (will the plastic flow without creating weak spots or cosmetic flaws, will the part pop out when done, etc). It’s by far the most expensive part of making Pickup, and we’re on a tight budget. But it’s nice having a manufacturing partner with the same business hours to answer your questions quickly.

That said, I’m not going to sacrifice speed for quality. I think I’m a week or two away from freezing the design, but I’ve been printing models like crazy to fine tune fit and bits of functionality. 

Plastic Part Design for Injection Molding: An Introduction, Robert A. Malloy

I’ve also been digging into this thrilling book, and I’m such a nerd about this that I have to stop myself from reading beyond my immediate need. But if you’re ever pondering injection molded parts, I’m here for you.


Last update I mentioned that Jeremy ordered prototype boards. These have been painfully hand-assembled and booted up.

front and back of populated Pickup proto board

He’s working through bringing up all the I/O now. Here’s the CO2/rH sensor producing its first raw data:

CO2 sensor board, Pickup and debugger board with screenshot of CO2 readings

And here’s checking data from the spectrometer:

sensor board in green light, screenshot of raw spectra values

And with both mechanicals and electronics getting close to done, we can bring them together to test rigidity of the assembly and that everything lines up.

Jeremy putting prototypes together.
proto board in proto case, lit up


Jeremy’s continuing to crank through sensor modules and we’ve been wrangling with the little problems that come up when you take things from prototype to production. I’ve also been sourcing lots of little parts like connectors and cables, and I’m going to be ordering them.

I’ll be ordering tooling and getting back to software. There, it’ll feel like progress is coming faster and we’ll have more frequent updates. Thank you for your patience!

Crossposting from Pickup Kickstarter update #3