DIY High temp thermistor probe

I've been toying around with the idea of creating a temperature control unit for a bullet style barbecue smoker using an Arduino, a thermistor and some 5v fans. The idea is to make something similar to what they sell at pitmasteriq.com and other places, only cheaper. Due to the cost constraints I'd like to use a thermistor instead of a thermocouple. I can't find any reasonably priced thermistor probes that will get me into the 2-300 degrees C range. My question is this: is it possible to buy a glass bead thermistor and enclose it in copper tubing and still get an accurate reading? The thermistor I have in mind is Digi-Key Part Number 495-2135-ND and I'm thinking of encasing it in either a plumbing stub out, or some kind of copper pipe with one end braised or pinched shut and then filled with a high temperature silicone sealant. The environment that the probe will be in is likely to never be much above 200* C, but it will be quite humid. Is this possible?

You have the problem of how to attach the wires to the thermister - solder won't work at 300C.

My thought on the joint was to use a lineman's splice or something similar, and let the filler in the probe body help keep the joint from moving too much. Sounds great in theory but how it would hold up in practice is another story. If the question of solder can be solved, is the rest of the idea plausible?

Thermocouple too expensive ? Come again...
http://goo.gl/5ZXld

// Per.

Thermocouples themselves aren't the cost issue. Finding them cheap isn't difficult. It's the roughly $18USD amplifier setup that's the cost issue.

My question is this: is it possible to buy a glass bead thermistor and enclose it in copper tubing and still get an accurate reading? The thermistor I have in mind is Digi-Key Part Number 495-2135-ND and I'm thinking of encasing it in either a plumbing stub out, or some kind of copper pipe with one end braised or pinched shut and then filled with a high temperature silicone sealant.

That is the way most temperature sensors are used in industrial applications, (typicaly J or K thermocouples or RTD based sensors), using a so called 'thermowell', that isolates the sensor from the medium to be measured while still accurately reflecting the temperature of the medium. No need to fill the the well with any liquid (in fact that is a safety issue in itself), just having the sensor tip mechanically in contact with the far inside end of the thermowell works just fine. The use of a thermowell also allows the sensor to be removed for testing or repair without needing to shutdown the industrial process to remove the sensor safely.

The environment that the probe will be in is likely to never be much above 200* C, but it will be quite humid. Is this possible?
Given time the inside end temperature of the thermowell is the same temperature as the medium, that is the whole point of using one.
Be sure any sensor type you use is rated to well above the maximum temperature the medium can reach, and that includes any lead wire of the sensor inside the thermowell, which is usually the limiting factor.
Lefty

Thanks for the insight retrolefty. I'm not familiar with the more industrial aspects of things, so that all helps a lot. I'll give it a go and see how things turn out.

thedave87:
Thermocouples themselves aren't the cost issue. Finding them cheap isn't difficult. It's the roughly $18USD amplifier setup that's the cost issue.

Get a couple of free samples from Maxim of a MAX31855 and a cheap SMD-DIP adaptor and you're in business.

thedave87:
Thermocouples themselves aren't the cost issue. Finding them cheap isn't difficult. It's the roughly $18USD amplifier setup that's the cost issue.

Just because everybody uses the "easy" way and uses a very expensive Thermocouple amp with SPI interface, you can do without, and if the other option is a NTC resistor, which isn't particularly precise...

Have a look here: Building LM358 Thermocouple Amplifier Demo

// Per.