Beginner questions on temperature measurement

Hi,

We are working on a temperature measurement project. The goal is to have fun, but it might turn out to be practical.

The 'target' to be measured will be grape buds and grape leaves experiencing frost. Temperature readings will be taken outdoors just before dawn, when air temps will be around 0 C (+/- 5 C). The lowest expected surface temperature will be around -5 C. The results will then be compared to plant tissue damage (or lack of damage) observed over the next few days. Eventually, we would like to put 50-100 sensors in the vineyard and hook them up with wireless units.

Here is our prototype. It is using two thermistors. The '3 wire' digital thermistor is visible in the bottom right. Two wire thermistor leads can be seen in the upper left.

We have logged data at 5 second intervals to both a file, and command-line window of a Windows-XP computer. The results looks accurate.

The first problem is calibration. For a growing grape bud, the difference between -2 C and -4 C can be life or death, so +/- 0.5 C accuracy is critical. It seems an ice water bath will provide a good 0 C 'known' temperature value for calibration in the field. Platnium sensors seem the best for stable, high resolution sensors. For now, we are Murata Electronics NXFT15XH103FA2B100 (2 wire), and Maxim Integrated DS18B20+ sensors. We have 3 to 5 sensors of each type. It isn't clear how to water-proof them for testing the 'ice bath' procedure.

Second, we would like to have an accurate hand-held unit for spot checks. When we look at handheld units for sale, we don't understand enough to know what we are looking at. Apparently, a PT-100 RTD is recommended, but PT-100 seems to be a generic title for a vast number of sensors, none of which are called PT100. So, if we get a platinum PT100 sensor for surface readings in the 0 C range (high quality), it isn't clear what hand-held device we need for data-logging and real-time display. Further it seems we can hook a PT100 sensor directly to our Arduino and (if it can operate at 0 C) take our measurements with the prototype set-up. I'm not sure we encounter the water-proofing issue with a platinum PT-100 RTD. I can't make sense of the parts listings.

Any suggestions would be appreciated.

JoeBob

JoeBob:
it might turn out to be practical.

It is using two thermistors. The '3 wire' digital thermistor is visible in the bottom right. T

The first problem is calibration.
we are Murata Electronics NXFT15XH103FA2B100 (2 wire),

and Maxim Integrated DS18B20+ sensors.

a PT-100 RTD is recommended, but PT-100 seems to be a generic title for a vast number of sensors,

I think there are some you haven't mentioned but I submit you would be better off attending to one type of sensor.

Even though you already have thermistors working, I would suggest you go for the DS18B20. They are cheap, need no calibration, accurate to a fraction of a degree, simple to install using a one-wire sytem, and are available in a weatherproof package. They are also popular and well-supported in this forum.

Your second problem is far from clear but, whatever it is, an Arduino can handle it. Quite possibly the same Arduino.

The DS18B20 is available in a form that has a stainless steel cover and waterproof sheath.

FYI the accuracy out of the box is +/- .5C,but the drift is only +/-.2C within 1000 hours. In other words, if you can calibrate it with a more accurate thermometer then you have a +/-.2C accurate device and can count on that accuracy for over a month.

The DS18B20 is available in a form that has a stainless steel cover and waterproof sheath.

I am concerned that the DS18B20 case makes it unsuitable for taking surface temperatures. We will be touching part of a plant. The air temp will be about -1 C, the plant surface will be about -3 C..

I guess we should simply try out both types of sensors and see what happens. I don't think I would trust our sensors in a water bath, but on an icy morning, things should be dry enough for taking surface temps.

JoeBob

You might consider looking at infrared thermometers which would more accurately give you the surface temperature of the grapes. The moment you butt a plastic DS18B20 to a grape you insulate it, and if you use a stainless steel one you heatsink the grape.

Lots of information at publiclab.org. 🎈 Public Lab: Circuit Diagrams for 5V Melexis Sensor