This incubator is very much a result of trial and error, not enough time, and a self-imposed restriction to build it out of as much garbage/found goods as possible. However, it was put together for ~$350 and it actually works!
Even though the incubator sustains life, it is far from perfect and it is worth looking at possible improvements. Below is a non-exhaustive list of some potential changes:
Code/Circuits can be more efficiently implemented. Im hoping others in the community might have time to improve the implementation and will share it. The current code was put together by brute force and is quick and dirty. In particular, potentially employing threading or protothreading in the Arduino code might be fun to try out. Please let us know if you make use of these plans/code.
Find a cheaper CO2 sensor or coming up with a DIY version. This needs some thought, but starting with the components required for an NDIR sensor is a place to start. The sensor used in this project is by far the most expensive part and represents about ~60-70% of the total project cost. Cheap CO2 sensors would not only decrease the cost, but would also allow for multi-point measurements around the incubator volume which could be used to ensure a more uniform CO2 concentration.
Optimizing CO2 stability over time. Right now, when the solenoid opens CO2 (at ~100%) enters the incubator. This results in the CO2 content quickly ramping up to the setpoint. Although the time-averaged CO2 content from this setup is in the range of 5.0 ± 0.2%, the level can slowly fluctuate due to the stepping mechanism. There are a variety of ways to solve this, including: pre-mixing the CO2 with atmosphere prior to pumping into the incubator, finer control over the input pressure/flow rate, faster gas mixing, experimenting with incubator size, improving air tightness, etc.
Possible implementation of the Arduino PID feedback library to control CO2 and Temperature. PID might be worth examining for finer and more stable temperature/CO2 control. PID/PWM control was briefly tested but not in great detail.
There are lots of alternative items one can use for the incubator chamber. There may be some better choices out there. Would be kind of cool to have windows.
Integration with DIY microscopes to allow for monitoring of cell growth/dynamics.
Twitter enabled. Get status updates, send control messages, control any DIY microscopes etc. Send temp/CO2 data to thingspeak.
For point #6 & #7, Raspberry Pi control might be a better option with a little more flexibility and internet integration.
When the heaters are on the incubator draws ~65W (give or take). Camping solar panels that supply 80W at 12V might be really cool for cell culture in the field and/or remote/resource-scarce regions.
Better ports for passing wires into and out of the incubator.
Devise of a better mechanism for latching the door closed.
Alarms and/or LEDs to indicate low levels or the activation of heaters/solenoids.
Thats it for now. Please let us know if you build your own!
>>> Design and Build
>>> Arduino and Circuits
>>> Proof is in the Pudding
>>> Possible Improvements