ADC0831 and thermocouples
peleus21
Posts: 16
has anyone used a A/D Converter like the ADC0831 or ADC0838 to convert one or more thermocouples - the key is more then one thermocouple - for the basic stamp?
i am going to try this but was wondering if anyone else has and what their levels of success was.
i am going to try this but was wondering if anyone else has and what their levels of success was.
Comments
As I've learned from Tracy Allen, there are chips for reading thermocouples directly that make the whole process easier.
I've used this thermocouple amplifier with a MCP3002 ADC chip successfully. I had trouble when I tried using the same amplifier with a MCP3208 chip. I think the MCP3208 needs a lower impedance source than the MCP3002 (my guess since it work with one ADC and not the other).
SparkFun sells a chip to directly interface with a thermocouple. I think this one is similar to the one Tracy recommended (you can find Tracy's recommendations in thread linked to previously).
You need to make sure whichever amplifier you use matches the thermocouple type you're using.
I don't know enough about op-amps to know if you could use one in place of a thermocouple amp. I'd imagine you could. The thermocouple amp has additional features (cold junction) which would likely make them more accurate than a regular op-amp.
Did you look at the $12 chip that doesn't require an additional ADC chip? SparkFun also sells type-k thermocouples for $14.
Are you sure you need thermocouples for your application? If you described what you're trying to do in more detail, someone may have a less expensive alternative than using thermocouples.
I often use a multi-channel 16-bit ADC with a full scale range of 0.256V to interface directly to thermocouples. The least significant bit is then around 7 microvolts, so there is good resolution.
You can amplify the thermocouple output with an op-amp and apply the output of that to a lower resolution ADC. The LM358 would work, but you have to allow for the fact that it has an offset voltage typically 2mV, which amounts to a possible 50°C error. That could be nulled out by calibration. Also the offset will drift by ~10 µV/°C. There are op-amps with much lower offset voltages and much lower offset voltage drifts, but more $$ of course.
Either approach, the op-amp, or the high resolution converter, requires a separate temperature reference and some math. The procedure is documented for both the Stamp and the Prop, but no bones about it, it is extra programming effort.
The chips like the AD595 or the MAX6675 contain the reference temperature compensation and amplifier, and that is how they pay for themselves.
Thermocouples produce a very low voltage and as such easily picks up extraneous noise. You have to take care to avoid that and to keep the noise from messing up the readings.
+1... can you use thermistors?
i am not familiar with the difference though.emf will be a big problem in my project. it uses a water manifold and pumps to do stage heating of my house. the sensors would be taped to pipes to monitor heat in pipes. pumps are 24 volts a/c.
the reason for thermocouples or thermistors (if they would work) is to get reading from pipe only - some of which i have found have a fitting to take a reading right from fluid inside pipe. i also have to monitor pumps and water flow. this is the other reason for the thermocouple/thermistor. the 8 channel MCP3208 uses 4 inputs to read 8 sensors, and the relay board uses another 4. this allows me to read and control 16 sensors and 8 relays from 8 input/outputs.
Thermocouple wire is rather expensive by the meter, and both thermocouples and thermistors don't fair too well out on long cables in the presence of EMI. There are IC sensors you can tape to a pipe or insert into a thermowell, such as the LM34 or the AD592, and they produce a high level voltage or current. There are also digital temperature sensors such as the DS18B20 that can be strung on one wire pair like holiday lights. Finally, if the console is located far from a cluster of sensor locations, there is a possibility of transmitting the readings wirelessly via XBee.
i am using the stamp to control for now as i am still learning programming.i do have a prop development board to experiment with though. i bought some lm358 chips and some thermistors to experiment with. i also built a board that would let me interchange chips - the adc and amps as long as connections stayed the same and can bypass the amps by using a set of jumpers i am calling it my "sensor board". my sensors are going to be a maximum of 5 feet from my "sensor board"