TempDoctor Reflow Oven Controller

This is one project that I actually intended to put on kickstarter / indiegogo / crowdsupply, but after I had developed a prototype there was an unexpected shift in my career and now I don’t have the time to pursue this project anymore. I am just publishing my work and perhaps someone else could take the challenge to make it into a crowdfunding.

When I needed a reflow oven for my lab, there was nothing on the market that I liked. Without spending a couple of grands on a professional convection oven for prototyping and small batch assembly, the only available stuff was DIY kits that did not satisfy my performance requirements. Since I was sure I can make a good performance DIY kit with a sleek design which would cost only a few hundred bucks, I set to work.

TempDoctor module

The PCB was designed to fit an already existing case, as custom molding is too expensive for low-volume production. I found some kind of cheap mouse to be a good source of nice round and slim cases that can take a PCB as front panel (details in hardware files).

 

Convection oven integration

http://steba.com/product/baking/bake-oven-kb-14/?lang=en

The solid state relay simply replaces the original electromechanical relay and the controller is attached in place of the original temperature control knob.

I added extra thermal insulation with fiberglass sheets and aluminium tape. This particular configuration is able to ramp up with up to 1.5°C/s.

Functionality

• Superior control with fast set point stabilization and no overshoot
• Automatic parameter learning by step response analysis
• Manual setpoint with simple timer
• Automatic profiles – download via USB and save to memory (up to 8 profiles)
• Intuitive, fast and responsive operation with touch buttons
• Main display shows actual temperature, target temperature, rate of change in actual temperature, heater status (on/off), operation mode and timer count

 

Source code and hardware files

https://gitlab.com/dorinclisu/oven-controller

The temperature control is not based on PID because this method leads to very poor performance in this application (large thermal inertia and delayed response).

I developed a control algorithm based on a first order approximation of the differential equations describing the thermodynamics of the oven, and it is fairly close to optimal control.