Reboot

TL;DR: The project has been on hiatus for a while, mainly due to very bad luck. We restarted work only a few months ago: the electronics of the piezo driver have been reproduced, and the first mechanical assembly is on the way.

Things don't usually go as planned, and basic assumptions can fail. This is a project update about the difficulties we came across while making, and the solutions we just found for them.

This project built upon an earlier micropipette project, driven by heavy steppers, and funded by the Gathering for Open Science Hardware (GOSH) in 2021. At the time, our workshop had access to a CNC lathe, which provided the tight tolerances we needed to manufacture a critical component: the tip-holder. This part has two important surfaces for sealing: one with the tip, and one with the shaft.

We needed the accuracy but, after a few prototypes, the lathe was no more.

Dead CNC lathes: 1

 

Dead electronics, no spare parts, and no funds for the expensive fix. RIP.

The broken lathe put the entire project on hold for years until 2023, when we finally gained access to a lathe in another of the university's workshops.

The 2-year old machine had not been used at all, but we managed to get it up and running (after some bureaucracy), and started making tip-holders.

Perhaps ironically, the first lathe we had access to had been donated to the university several years ago, and had also been collecting dust until Gastón got it working. History repeats itself.

 

One tip-holder in the making.

In pararllel, and due only to good luck, we shared our ideas with Renan Soares, from Universidade Federal do Rio Grande do Sul. I had learnt the basics of automation from him, during the reGOSH 2019 residency at Porto Alegre, and had been looking forward to working together for years. This project arrived just in time: he also wanted to develop open-source piezoelectric actuators with closed-loop control. We finally got a chance to collaborate! and got to it immediately.

After a lot of research and some logistics, Renan reproduced the basic principle that will drive the miniaturized pipette actuator. The next major milestone will be integrating an encoder, precise enough for the volmetric accuracy needed for liquid handling.

 

A single channel actuator, ready for testing and charachterization.

Remember I just wrote that history repeats itself?

As if that were a hypothesis in need of verification, the new lathe also failed after a few months. It was new, but it had sat unused long enough for the warranty to expire, making the fix is also too expensive for the university's budget (which has suffered monstruous cuts this year, btw).

Dead CNC lathes: 2

I feel I should clarify that these failures were electronic. There is no way that we caused them, but we still felt... jinxed.

We are excited to develop tools for top-line automation in biology, and to work with Edwin and Renan. But this kind of bad luck is frankly devastating.

What could we do? Well, here is where watching dozens of hobbyist electronics videos on YouTube paid off: enter PCBWay's machining service.

 

Tip-holder and stainless steel shaft, all custom, in one week.

The service is a lifeboat: we uploaded the design files with some drawings, and got shiny parts in two weeks. While definitely not as cheap as in-house parts, and also far from being local, it means that development can continue.

In the meantime, the robot that will one day use the tool for science has seen great updates, and is now pipetting yeastly microscopy experiments with ease.

Everything we do is open-source, and you're welcome to join our discord and watch the repos at GitLab.