RES31A10DDFR datasheet(21/50 Pages) TI2

21 / 50 page

7.3.2 Ratiometric Drift

The ratiometric matching of the RES31A provides a benefit not just for initial conditions, but also when

considering parametric drift. The resistors must be considered individually, in absolute terms, and ratiometrically

to each other, in matched terms. The absolute temperature coefficients of each resistor show strong correlation,

with the coefficient of RIN1 comparable to that of RIN2 and the coefficient of RG1 comparable to that of RG2.

The absolute temperature coefficient (in Ω/°C) of each RG is approximately Gnom times greater than that of

the comparable RIN; therefore, the normalized absolute temperature coefficient (in ppm/°C) of every resistor is

approximately the same.

The resistors of the RES31A are interdigitated, and occupy a small footprint; thus, the die temperature of

the device is effectively common to each of the four resistors. As the temperature changes, each resistor

experiences a similar temperature rise. The resistors have very similar temperature coefficients; therefore,

the ratio of RG to RIN is well preserved. For example, the RES31A40 has a typical absolute temperature

coefficient of approximately 12ppm/°C for RIN or RG. When considered in ratiometric terms, the typical

temperature coefficient of tD1 or tD2 is ±0.2ppm/°C, and the temperature coefficient of tM is ±0.05ppm/°C.

Ambient temperature, humidity, heatsinking, board cleanliness, and other related factors can impact the settling

time of the RES31A, so validation testing is performed in a low-humidity environment with rigorous board

cleaning procedures.

7.3.3 Ultra-Low Noise

Noise in resistors can be evaluated in two separate regions: low-frequency flicker noise and wideband thermal

noise. Flicker, or 1/f noise, is extremely important for systems that require signal gain at frequencies less

than 100Hz. The flicker noise for thin-film resistors, including the RES31A, is lower than that of thick-film

resistor processes. Thermal noise typically dominates in the region greater than 1kHz, and increases as resistor

magnitude increases. Noise is modeled as a voltage source in series with the resistor.

For a resistive divider such as the RES31A, the thermal noise as measured at the center tap of two resistors RIN

and RG is equivalent to the thermal noise of a resistor with value RIN || RG:

eN= 4kBTR

(16)

where:

• eN is the thermal noise density in nV/√Hz

• T is the absolute temperature in kelvins (K)

• kB is the Boltzmann constant, 1.381 × 10-23J/K

• R = RIN || RG

As an example, for the RES31A40 at 25°C:

eN= 4kBTR = 4×1.38E−23JK×278K× 100kΩ ∥ 400kΩ =35.1nV/ Hz

(17)

7.4 Device Functional Modes

The RES31A is typically used with two independently biased resistor dividers. RIN1 and RG1 in series form a

resistive divider, with RIN2 and RG2 in series forming another divider. However, the two dividers do not have to be

used independently. The resistors can be connected in series or in parallel like any other resistor.

Use one of the two GND pins to bias the part substrate. Connect the substrate to signal ground or a similar

low-impedance bias point or plane for best noise rejection. While two GND/SUB connection pins are available on

the device, connect only one of these to the ground plane. The two GND pins are internally connected through

the substrate, which is not intended to conduct significant currents. Connect only one GND pin at a time and

leave the other pin floating to prevent current return paths from developing through the substrate.

www.ti.com

RES31A

SLVSKC6 – DECEMBER 2025

Submit Document Feedback

Product Folder Links: RES31A

RES31A10DDFR 数据表(PDF) 21 Page - Texas Instruments

RES31A10DDFR 数据表(HTML) 21 Page - Texas Instruments

Html Pages

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

数据表下载