Proportional controller performance

[DurandA]

Hi @UnexpectedMaker and thanks for your open-source projects and YouTube videos.

I am making a pretty similar project for a hot plate where I replace the original controller board with a custom one. Since there is just space for a segment display and some buttons, reflow profile will be configured from a smartphone using BLE. The hardware is mostly final and I am now working on the firmware using ESP-IDF.

I stumbled upon your project and found that you use the same thermocouple chip (MAX31855) than me and I am wondering how did you settle on the current temperature controller. I started working on a PID controller but saw that that you are only using a differential input:

ReflowMaster/Code/Reflow_Master_v2/Reflow_Master_v2.ino

Lines 997 to 998 in 29f8a3f

	duty = base + ( 172 * perc );
	duty = constrain( duty, 0, 256 );

I did not try to tune the PID controller on the plate yet but was wondering how did you settle on a proportional-only controller? Does the temperature closely match the chosen profile? I also saw some recommendations to rather use bang-bang (on-off) controllers for heaters due to the large inertia.

[UnexpectedMaker]

I don't think PID is a good fit for reflow, so I implemented my own solution. It's not really differential control, and more a combo of a few techniques, that allows users to tune in a more understandable way.
But this really isn't a RM issue, so maybe my discord server is a better place to talk about it 👍
https://discord.com/invite/xAHpApP

[DurandA]

Whoops, I meant proportional controller.

Sorry for opening an issue and thanks for the Discord link!