I only have access to a MAX6675 amplifier for a type k thermocouple. I would like to measure temperatures below zero. Is this in anyway possible? If not, any suggestions to another amplifier I can get quickly that can do the job?
Browse here
Click on the operating temperature range you'd like.
That looks like a good site but I guess I should mention my task requires temperatures to be read around -80 deg C. Those seem to only go to -55 deg C.
1 Like
Knowing -80C would hve helped.
PM to JohnWasser, he helped some folks recently with weather balloon that needed -100C.
The MAX31855 is a new version of the 6674/6675. It is a 3.3V part, but it is pin compatible with the 6674/6675, has versions for T-couples other than K, and can do low temperatures (assuming the MAX31855 is kept above -20C or so). The datasheet says:
This converter resolves temperatures to 0.25C, allows readings as high as +1800C and as low as -270C, and exhibits thermo- couple accuracy of ±2C for temperatures ranging from -200C to +700C for K-type thermocouples.
IIRC, a K-type isn't as suitable for cold temperatures as some other types (like T?), but they are easier to get as they're the most common type. I think I've got a hobo datalogger around that has a T-type on it.
-j
I've never used a thermocouple before so all I could offer to the balloon folks was:
"So a thermocouple should be able to measure the desired temperature range but I have not worked with them before so there will be a bit of a learning curve. The various compensated thermocouple amplifier chips seem to have a low temperature limit in the -40°C range so I can't go the easy way. I guess what I'd have to do is hook the thermocouple to an op-amp to bring the -4 mV reading (120°C below ambient) up to about 4V. Since the measurements are relative to the "cold junction" ambient I probably don't need to worry about voltages above 0 so I can just use the op-amp as an inverting amplifier with a gain of 1000. That will give a resolution of about 3.9 microvolts which is better than 1/10°C resolution at the 'hot' end (0°C) and better than 1/4°C resolution at the 'cold' end (-100°C)."
Looks like the MAX31855 is an excellent choice. The K-type and N-type versions are accurate +/-2°C over the -200°C to +700°C range. The T-type version is accurate +/-2°C over the -250°C to +400°C range. You just have to keep the ambient temperature (temperature of the chip) between -20 and 85°C.
I'm using this MAX31855 , and it works perfect, but only on its own. If I try to combine him with a SD datalogger it stops working. Both uses the SPI bus. The SD , selectpin 10 and the Max with selectpin 9.
With the standard SD-Library coming with Arduino 1.0, I tried several MAX31855 Libraries, no one succesfull.
What could be a solution?
Some libraries for SPI devices assume there is only one SPI device. They initialize the SPI parameters (clock divider, bit order, mode) and assume they remain unchanged. Other libraries, such as the SD library, change the SPI parameters on each call to make sure they are set correctly for that particular SPI device. These two kinds of libraries won't work well together. You can make the second type of library work with the first type by re-setting the SPI parameters before each call.