My task was to measure water surface temperature and at 16 equal points directly below the surface to a depth of 4mm, and the air just above.
I could not just use one sensor and move it up and down because it would mix the fluid and change the results.
Iteration 1: Thermistors
Analog Devices has a kit with a "Linduino", and a LTC2983 24-bit digital temperature measuring system.
I purchased the kit and 16 thermistors, soldered them up and here's V1 (each grid square is 1cm)
This setup worked OK in air, but went crazy in water. I could seal the whole thing in epoxy, but had another idea...
Iteration 2: Thermocouples:
I had a bunch of K-type thermocouples, so I pushed them through a header form after taking all the pins out and tried it out.
The Linduino wasn't giving me stable values, so I tried it out on a National Instruments DAQ system.
The NI9213 is a 24-bit 16-channel thermocouple input module with differential inputs,
the cDAQ‑9171 is a bus‑powered, CompactDAQ USB chassis for the NI9213.
The sensor needed to be smaller, and the DAQ unit needed to be more portable and battery operated
A buoy with a laptop and CAT5 tether was kinda bulky also. We needed a better solution.
Iteration 3: smaller thermocouples:
I found some circuit board with 1.25mm spacing and put some 40AWG k-type thermocouples through the holes.
Then I wrapped between each thermocouple with sewing thread to build up and separate the individual sensors.
Then I epoxied the device very carefully.
I changed out the NI USB DAQ to a NI cDAQ-9191 WiFi DAQ that could be battery operated.
The sensor array was mounted at 15deg to a pair of floats so that half of the sensors would be in the water, and half out of the water.
Here's the LabView Block Diagram:
The data was interesting. When the probe was out of a mixed layer of water then put into the water, the graph was as expected.
When the probe was submerged completely and then raised half-way the data was a bit unexpected.
Instead of showing warmer air temperature, it showed cooler than water temperature due to evaporation cooling.
We found out that we couldn't tell if some of the thermocouples were in colder water, or just out of the water with evaporation cooling.
Of course the first time we put the probe in a real lake, a dragonfly immediately landed on it and caused all sorts of problems, after that, the minnows were nibbling on the probe.
No matter how hard we tried, a tiny wind would cause a wave big enough to submerge the entire probe.
Back to the drawing boards........
Top Comments