Overview
This project implemented the previously made RTD sensor to a bang bang control scenario. The plant chosen was a bowl of oil, and was to be dynamically heated using a soldering iron.
Initial premise
The premise of this project was to demonstrate correct bang control of a given plant, with a heating element and a cooling one. Now, It is difficult to directly interface heating to electrical control, because the heating element needs to be actuated by the mcu (arduino). So, I settled to use an old soldering iron to heat the plant, as it can be easily controlled with a relay.
Next, was choice of media. The obvious choice of water had a few issues: it would short out the soldering iron, and it would make a very slow unobservable response time. So, I settled to use cooking oil as it will give us a nice visible response time and will have no risk of electrical damage in case of leakage.
Implementation of plant
To safely connect the soldering iron to media, I needed a bowl that was cheap and modifiable, but wont leak nor burn from the iron. So I settled for a plastic container, and I modified the side to attach the soldering iron. To never worry about the iron burning a hole in the plant, I used an old ceramic tile at the bottom of the container.
Similarly, the fan was attached to the side of the container. The fan was driven with a lm9110H bridge: overkill for this purpose as directionality isn’t needed, but convenient.
Demo Video
Watch the project in action on YouTube.
Results
The main issue I faced was the Relay coil energizing disrupting power flow to the h bridge and rtd, causing oscillation off mis measurement. essentially, the temperature would suddenly drop to 0 during transition, exciting the coil again, and so on. To address this, I implemented decoupling capacitors at the coil and the rtd.