|Activities:||Test construction, CO2 sensor interfacing, pressure sensor interfacing|
|Technologies:||STM32 Cortex M0, CANopen, I²C|
Some research project called for the ability of measuring CO2 concentration under high pressure conditions, specifically up to about 70bar. A first test was performed using some 40mm copper pipe and fittings to validate the basics of the theory at 'shop-air' pressures (~7bar/700kpa). The result of this first test looked promising, so we scaled up for the building of a bit more serious pressure chamber to support full pressure and beyond.
In order to accomodate the sensor under high pressure, we have a high pressure chamber made from solid steel by one of our customers. This was modelled to be strong enough to survive beyond the 550bar safety release of the valve. The chamber is combined with a low-cost water cooled 300bar airsoft compressor.
We used a modified CANopen ambient sensor for temperature monitoring and to interface over I²C with the CO2 sensor module and also with a differential ADC breakout to connect with a high pressure sensor.
Low-cost CO2 sensors do not really measure PPM-levels, but rather perform some kind of parts-per-volume measurement. This means the appearent concentration will rise when the air gets compresed. Also, this is why these sensors have commands for setting the ambient pressure. Anyway, a practical problems arose from actual CO2 concentration of air indoors. The theory was based on a 400ppm concentration, which should give a reading around 28000ppm at 7Mpa. In reality, saturation occured much earlier. That's because indoors, this nominal CO2 concentration easily gets 2 to 3 times higher. Since the sensor is limited to 40000ppm reading, this meant the readings saturated at 2-3× less pressure than expected.
Test chamber specifications:
- Working pressure 30Mpa
- Burst pressure > 55Mpa
- Test volume ⌀ 50mm × 120mm = 235cm³
- High pressure feed-through for 7 wires
- Dimensions ⌀ 85mm × 220mm (without valve and connections)
- Weight ~6kg