Temperature Indicator Calibration

The temperature indicator output is based on the ideal junction temperature formula below. However, the results can vary slightly from device to device. This is due to the device’s transistor voltage threshold, Vt , which will vary within production allowances. The temperature indicator doesn't have a built in calibration mechanism so external testing may be required if the tolerances are outside the specific applications needs. Typically the temperature indicator works best for single trip points or over-temp applications. For more accurate temperature measurements an external temperature sensor such as the MCP9700 or MCP9800 can deliver the accuracy over a larger range of temperatures.

The ideal voltage of a single junction temperature can be calculated per the equation below.

(1)
\begin{equation} Vt = 0.659 – (Temperature Degrees C + 40) * (0.00132) \end{equation}

Some applications can take calibration data during the manufacturing process. For those cases, the temperature indicator can be measured and a calibration value can be calculated and stored in EEPROM or Flash Memory to offset the results of the ideal temperature ADC result equation. This may require single point or dual point calibration where multiple devices are measured. For example, a sample of 12 devices could offer a range of temperature measurements similar to the chart below.

TEMPERATURE DATA FROM 12 SAMPLE DEVICES (2)
\begin{equation} ADCResult = ((mode*Vt)/VDD) * (2^n-1) \end{equation}

mode (High Mode) = 4
mode (Low Mode) = 2
where n is the resolution of the ADC; 8 or 10 bits.

### SINGLE-POINT CALIBRATION

Single point calibration is done by measuring the device reaction at a single fixed temperature. This can offer an offset value to give a more accurate measurement for a trip point. But variation from device to device may affect any measurement accuracy at higher or lower temperature points around the test point. Therefore, single point calibration is recommended when the temperature indicator is used for a threshold or trip point based on the temperature of the device. Single point calibration can be expressed per the equation below.

(3)

With a sample PIC16F1937 device under the following conditions:
• Powered at 5V
• High-range 4Vt operation
• 25°C forced temperature

The measured Analog-to-Digital conversion gives a result of 561 decimal.

Typical Analog-to-Digital conversion result at 25°C is calculated as 554 decimal using equation (2) above.
For single-point calibration, the difference between the conversion result and the ideal A/D conversion result value is the calibration value.

(4)
\begin{equation} Calibration Value = ideal – measured \end{equation}
(5)
\begin{equation} 554 - 561 = 7 \end{equation}

Consequently, for this device the calibration value would be 7. This value can be stored in nonvolatile program or data EEPROM memory within the device for use when taking temperature measurements.

### TWO-POINT CALIBRATION

Calibration is performed at two temperatures from which we can determine calibrated temperature. As a result, this method is more accurate across a larger range of temperatures, but requires two distinctively different temperature calibration measurement points. Those calibration points can be used to develop the parameters of the equation below.

(6)
\begin{equation} ADC Result calibrated = A + (B * ADC Result) \end{equation}
(7)
\begin{equation} A = (Ideal @ T1 – Ideal @ T2)/(Actual @ T1 – Actual @ T2) \end{equation}
(8)
\begin{equation} B = Actual @ T1 - (A * Ideal @ T1) \end{equation}

Where:
T1 calibration temperature 1
T2 calibration temperature 2

### Calibration Application Note

Note AN1333:
Refer to ""Use and Calibration of the Internal Temperature Indicator” (DS01333) for more details regarding the calibration process.