← Revision log
Dwg. no.
LOG-006
Rev
F
Date
Read
2 min

Pit-pyro: building the sensor board

The custom circuit boards for the thermal sensor arrived. I built one and read frames from it on the Feather.

On day two I designed a small circuit board to replace the hand-cut perfboard under pit-pyro’s MLX90640 sensor. The boards arrived today, along with the extra sensors, so I built one.

The board

The board is 20 mm across. The sensor goes on the front. The back has the two pull-up resistors, the two capacitors and four holes for the cable, labeled with the wire colors.

A small green circuit board held in calipers reading 20 mm, showing the sensor outline, four leg holes and the text pit-pyro
The front of the bare board.
The back of the bare board with labels GND blk, SCL yel, SDA blu, VDD red, two resistor outlines and two capacitor outlines
The back of the bare board.

The sensor dropped in from the front and all four legs came straight through their holes. It dry fits like a glove.

The back of the board with the sensor's four gold legs sticking up through their holes
Dry fit, seen from the back.

Building it

I only had one 4.7 kΩ resistor left, so the second pull-up is 5.6 kΩ. Soldering on this board is much easier than on the perfboard.

The 10 µF capacitor stands 13.2 mm off the board, and the pocket behind the board in the head is 10 mm deep. I forgot to check its height when designing the board. For now it’s soldered standing, with some leg showing so it can be folded over later.

A black 10 µF electrolytic capacitor held upright in its pads on the back of the board, much taller than the board is wide
The 10 µF capacitor.

I soldered the sensor in without pressing it flush, so it sits slightly tilted. I’ll reseat it with solder wick.

Wired and tested

The cable is a 150 mm STEMMA QT cable with the header pins cut off one end, soldered through the labeled holes.

The finished board with red, blue, yellow and black wires soldered under their labels, the cable running to a plug, and the Feather board in the background
The board wired to its cable, with the FeatherS3 behind it.

The meter showed no shorts between the wires, and both pull-ups read their values. The Feather needed a reflash before it would show up over USB. After that the firmware found the sensor at address 0x33 with no bad pixels.

I aimed it at a flat wall and saved three frames. They averaged about 26 °C, with a pixel spread of 0.36 °C. There are no stripes like on day one. There is a faint pattern every four columns, about 0.3 °C, that the perfboard sensor from day two didn’t show.

The design files are in hardware/sensor-carrier.

Next

The sensor has had enough attention for now. Next are desk tests for the parts I haven’t tried yet:

  • The 0.96” OLED screen.
  • The DS3231 clock.
  • The LiPo battery.
  • The trigger switch.
  • The menu button.
  • The aiming laser.
  • The buzzer.