FAQ (よく聞かれる質問)： MAX31855

No. While the MAX31855 does perform automatic cold-junction compensation, it does not performany linearization calculations. Linearizing the thermocouple results needs to be performed externallyusing the NIST calculations and coefficients at the NIST ITS-90 Thermocouple Database: http://srdata.nist.gov/its90/main/

As an alternative, the MAX31856 performs on-board linearization for all supported thermocouple types.

No. The MAX31855 does not have any on-board 50Hz/60Hz notch filtering, nor does it have any high-frequency filters. A 10nF ceramic capacitor (maximum value) should be placed as close to the MAX31855 package as possible between the T+ and T- pins. Additional 10nF decoupling capacitors (maximum value) should be placed from T+ to GND and from T- to GND. As implemented in the MAX31855 Evaluation kit, a small ferrite bead or series resistor (< 1000Ω) can be added to help withfiltering. However, for high-frequency noise (> 100MHz), the MAX31855's ADC will produce large offset readings.

As an alternative, the MAX31856 was designed to have 50Hz/60Hz notch filtering as well as EMI/RF noise filters.

The MAX31855 implements a read-only Mode 1 SPI interface.

No, the MAX31855's SPI interface must not have voltage higher than 4.0V applied. High voltage andhigh current (through a voltage divider) will induce self-heating resulting in cold-junction andthermocouple temperature errors, as well as possible damage to the SPI interface pins.

There was a trim issue with two lots of MAX31855KASA+ devices. These lots had a date code of 1352A2. The lot codes affected were 253AL and 253AK. No other lots affected were released for sale. Only the 253AL and 253AK lots of the MAX31855KASA+ with date code 1352A2 were affected by the mis-trim, which trimmed the devices for J-type thermocouples. These devices cannot be re-trimmed, so they must be returned to Maxim Integrated for replacement.

Yes, but you will lose some fault detection capability since you are intentionally grounding the T-portion of the cable, and you may experience a slight decrease in thermocouple temperature accuracy since you are bypassing the internal reference voltage. It should also be noted that if T- is grounded, it should be grounded to the same reference as the MAX31855 GND. If the thermocouple cable is connected to a different ground reference, the two "ground" potentials will likely differ, leading tolarger errors. You do not want to have any electrical contact at the thermocouple tip, so grounding thethermocouple tip, or having it connected to a grounded path is prohibited.

The MAX31855 uses its own die temperature as the cold junction temperature when compensating for the thermocouple value. The actual cold junction is located where the thermocouple cables contact the circuit board (or other connector). The farther away this contact is from the MAX31855, the more likely there is a difference in temperature between the actual cold junction and the MAX31855 die temperature. This temperature difference will lead to errors in the reported thermocouple temperaturevalue since the MAX31855 automatically uses its die temperature as the cold junction compensation value. The remedy is to make sure to keep the actual cold junction as close as possible to the MAX31855 so that they have as little difference in temperature as possible.