[JBC245] A simple and easy-to-use JBC245 soldering iron controller

I. Introduction
        Currently, there are many open-source projects for the T12 control board, but fewer projects support the JBC245. Firstly, due to the low internal resistance of the JBC245 soldering tip, these projects all rely on a dedicated 24V external power supply, resulting in a larger size and limited versatility. Secondly, most control boards use OLED screens to display information, increasing BOM costs and complicating the soldering iron's usage logic. This project addresses these pain points while also offering advantages such as small size, rapid heating, and low cost.
 
II. Effect Demonstration
 
Front view:
Back view:
Controller temperature recovery effect. The green line represents the PWM duty cycle, the black line represents the current, the blue line represents the target value (280℃), and the red line represents the actual value (real). During the experiment, the solder joints were continuously melted with the soldering iron. As you can see, the deviation of the target from the real is very small (within 3℃).
III. How to Replicate
        1. Making the Soldering Iron Handle
        The wiring diagram of the JBC245 soldering iron tip is as follows (JBC T245 pinout - Page 1 (eevblog.com)):
                 This project only includes the control board. The soldering iron handle can be purchased on Taobao and Xianyu. Generally, a good quality soldering iron handle can be found for around 30 RMB. Unlike other JBC open-source projects, the control board of this project requires a 4-wire soldering cable. Please note that it is 4-wire. Typical JBC open-source projects use 3-wire cables. The cable can also be purchased on Taobao. The keyword is "soldering station cable". I used a 1.2m long, 40mm outer diameter 4-core cable.
        The reason for using a 4-core cable is that this project adds an NTC thermistor inside the soldering iron handle to measure the cold junction temperature in order to improve temperature measurement accuracy. The actual temperature is the NTC cold junction temperature + thermocouple temperature. I believe this temperature measurement scheme is more reasonable. The reason is that the Seebeck effect of thermocouples only depends on the temperature difference; in other words, thermocouples can only measure the temperature difference, not the absolute temperature. Therefore, an NTC is needed at the end to measure the cold junction temperature. This project uses an NTC with a B value of 3950. As shown in the figure,
                    the NTC is placed inside the handle and connected across GND and RT. The NTC is not reversible.
        The thermocouple data for the soldering iron comes from JBC T245/C245 Thermocouple thermal coefficient - Page 1 (eevblog.com). I extracted the data from the figure, performed linear fitting, and used the slope × voltage to obtain the temperature difference. Superimposing this with the NTC at the cold junction gives the absolute temperature of the soldering iron tip. The advantage of linear fitting is that it is relatively stable when extrapolating. It also has simple parameters and consumes less MCU resources. The following figure shows the result of linear fitting and extrapolation to 400℃.
        2. Programming
        This project uses an STC8H1k28 as the main controller. P3.0 and P3.1 ports are used as programming ports. Due to space limitations (actually, we only realized the programming port wasn't drawn after we finished designing it), four wires need to be run from the solder
  
pads for programming (please don't criticize), as shown in the picture below. (The picture shows the first failed version; the programming port is unchanged.) After the wires are run, a USB-to-TTL module and a few test clips are used for programming. The official STC-ISP software is used for programming, with the internal IRC frequency set to 32MHz.
        3. Temperature Control Algorithm
        This soldering iron uses a variable-parameter PID algorithm for temperature control. When the temperature difference is greater than +20℃, it heats at full speed (mainly during cold starts). When the temperature difference is within 20℃, the PID controller takes over. The relationship between the proportional coefficient Kp, integral coefficient Ki, and error is as follows:
Using the function relationship shown in the figure above for temperature control has clear advantages. When the error is near 0, the system approximates ordinary PID control. When the error is large, Kp increases while Ki decreases. Increasing Kp allows the system to converge quickly, while decreasing Ki reduces the integral contribution when the error is large, thus suppressing temperature overshoot. Practical testing showed that this control method performs well. The convergence speed is faster than pure PID, and the overshoot is smaller. In this project, the D parameter was directly set to 0.1. I didn't spend too much time on precise parameter tuning, so there is still some room for optimization of the PID parameters. The relationship between Kp, Ki, and error is given below.
Finally, the MOS used in the project is AP90P03Q, with the same package as the one on the PCB. For some reason, there is no schematic diagram, so it can only be attached.