Improving the Response Time of the TMP118

Introduction

The response time of a temperature sensor can be a critical feature needed for applications such as smart watches, medical equipment, or gas meters. The benefits from a faster response time can range from allowing a system to make quicker adjustments to cooling, providing a more detailed range of data for a user’s sleep pattern in a smart watch, or alerting medical staff to sudden changes in a patient’s condition. The TMP118 comes in an ultra-small, ultra-thin PicoStar™ package, and the low thermal mass allows for competitive response times. This response time can be further improved through the implementation of a Flexible Printed Circuit (FPC) board.

Data

Figure 1 shows rigid and flex PCB response time data.

Figure 1 Rigid (Left) and Flex (Right) PCB Response Time Data

Thermal mass is a significant factor to consider when designing a system for speed and responsiveness. Not only does the IC need to have a small footprint, but for even greater response improvement, the circuit board needs to contain minimal thermal mass. Some possible ways to go about this are designing the sensor on a thinner FR4-based board with minimal layers; however, to maximize the capability of the TMP118, it is strongly recommended to consider an FPC board.

To better explain this, testing was conducted using a 62mil FR4-based prototype board and the TMP118EVM FPC sensor board (5.1mil) was conducted to collect data comparing the response time of each board when submerged into an oil bath running at 75°C. The TMP118 sensors on each board read an ambient temperature of nearly 25°C prior to submersion. In the testing for the Rigid (FR4-based) PCB, an average rise time of 1.4 seconds was observed. Meanwhile, the flex PCB was observed with an average rise time of 0.16 seconds, a nearly 87% improvement over the FR4-based board.

The significant improvement in the response time with the FPC Board is attributed to the low thermal mass, allowing the TMP118 sensor to quickly absorb the temperature of the environment (or surface) compared to a rigid board which takes longer due to the FR4 material absorbing some of the ambient heat in the environment, slowing the response time. The calculated thermal mass of the FPC board is approximately 0.769J/°C compared to 1.76J/°C for the Rigid board, over two times the mass of the FPC board. Both calculations included the thermal mass of the TMP118 device (0.14mJ/°C). The mass of these boards was found using the constants and equations in Table 1 and Equation 1.

Table 1 Board Dimensions and Constants

Board Dimensions (mm)
Rigid PCB Board	12.7 (l) × 12.7 (w) × 0.76 (h/thickness)
FPC Board	177.32 (l) × 2.5 (w) × 0.3 (h) at connector; 0.13 (h)

Board Constants
Density	Rigid PCB Board	FPC Board
	Copper: 8.96g/cm³
	FR4: 1.85g/cm³ Solder Resist: 1.4g/cm³	Polyimide: 1.43g/cm³
Heat Capacity	Copper: 0.385J/g°C
Heat Capacity	FR4: 0.85J/g°C Solder Resist: 0.9J/g°C	Polyimide: 1.1J/g°C

Thermal Mass Formula for Rigid (FR4-Based) Board and FPC Board

${M a s s}_{T o t a l} = {M a s s}_{C o p p e r} + {M a s s}_{S u b s t r a t e}$

${M a s s}_{C o p p e r} = L e n g t h o f T r a c e \times W i d t h o f T r a c e \times T h i c k n e s s o f C o p p e r \times D e n s i t y$

${M a s s}_{S u b s t r a t e} = A r e a \times T h i c k n e s s \times D e n s i t y$

$T h e r m a l M a s s = ({M a s s}_{S u b s t r a t e} \times C_{S u b s t r a t e} + {M a s s}_{C o p p e r} \times C_{C o p p e r}) + {M a s s}_{I C}$

Equation 1. ${M a s s}_{I C} ≅ 0.14 m J / ˚ C$

Note: Substrate = Board material such as Polyimide, FR4 (and so forth) and
C_Substrate = Heat Capacity of Material

The combination of a circuit board with low thermal mass and the TMP118 sensor can provide a meaningful and easy-to-implement answer to systems which require ultra-fast response times while maintaining accuracy.