What is the cause of Arduino glitching in this DIY soldering station?

Question

I was following this (Spanish/Spanglish warning) tutorial to create a DIY soldering station. I created the schematic, PCB and ordered the parts to assembly, used a cheapo 5 V and 24 V AC-DC power supplies to power the Arduino and heating element and everything seemed fine until I turned it on and noticed that my OLED display started to glitch when the 24 V begins to heat the tip.

My first suspect was EMI (which I have no idea on how to deal with) from the supply since I put everything in an old pc power supply enclosure, the PWM stays 100% so the tip heats without control and I have to restart the station. So my question is: What do you think is the cause?, I want general ideas because I know not everyone has time to check everything and I can research further.

I will add a general idea on how the attached Arduino code is working:

1. Starts and set up the PID control
2. Measures temp of the heating element
3. Calculates the PID depending on the set temperature and modifies the PWM signal which controls a MOSFET driving the 24 V for the tip
4. Refresh the display
5. Repeat 2 - 4

I'm using interrupts for the encoder and u8x8 for the OLED.

Code -> https://github.com/cocampo5/Soldering_iron_pid/blob/master/Soldering%20Iron/src/main.cpp

Ignore the big-a55 resistor. The 24 V supply at the beginning activates the short circuit protection and I thought a resistor as dummy load could help (it just created a ton of heat).

Answer 1

You don't appear to have any bypass capacitors. At a minimum, add a large electrolytic (20-200 uF) on the incoming 24 VDC supply, and film or ceramic capacitors (100-470 nF) on the output of the regulator, and the power supply pins of the Arduino. Also for the RESET line.

Answer 2

One of the main issues with this board specifically and 2 layer boards in general, is that long traces all over the place means more impedance, more crosstalk and higher chances to act as antennas, picking up radiated emissions. The ground traces in particular need to be as direct towards the GND pins of your supply connector as possible. Don't hesitate to bring them together when they meet. In this case the MOSFET ground in particular is delicate because that's probably where most of the current will pass.

As already mentioned, a complete lack of capacitors on a PCB pretty much always means it was not correctly designed. A bulk cap near the supply connector as well as decoupling caps near other supply pins.

Your power supply might be a culprit, there is not enough information to tell. It doesn't hurt to put a TVS on the 24V supply. Some 30-33V rated. I once had a crappy bench power supply which liked to give my PCBs a "horse kick" in the form of a nasty spike every time I flicked the power on.

Arduino in itself doesn't have the best reputation when it comes to EMI (or anything else). Neither does AVR. So there may very well be problems on that board.