The TMCS1100-Q1 application design procedure has two key design parameters: the sensitivity version chosen (A1-A4) and the reference voltage input. Further consideration of noise and integration with an ADC can be explored, but is beyond the scope of this application design example. The TMCS1100-Q1 transfer function is effectively a transimpedance with a variable offset set by V_REF, defined by Equation 28.

Equation 28.

Design of the sensing solution first focuses on maximizing the sensitivity of the device while maintaining linear measurement over the expected current input range. The linear output voltage range is constrained by the TMCS1100-Q1 linear swing to ground, Swing_GND, and swing to supply, Swing_VS. With the previous parameters, the maximum linear output voltage range is the range between V_OUT,max and V_OUT,min, as defined by Equation 29 and Equation 30.

Equation 29.

Equation 30.

For a bidirectional current-sensing application, a sufficient linear output voltage range is required from V_REF to both ground and the power supply. Design parameters for this example application are shown in Table 10-4 along with the calculated output range.

These design parameters result in a maximum linear output voltage swing of 4.65 V. To determine which sensitivity variant of the TMCS1100-Q1 most fully uses this linear range, calculate the maximum current range by Equation 31 for a unidirectional current (I_U,MAX), and Equation 32 for a bidirectional current (I_B,MAX).

Equation 31.

Equation 32.

where

• S_A<x> is the sensitivity of the relevant A1-A4 variant.

Table 10-5 shows such calculation for each gain variant of the TMCS1100-Q1 with the appropriate sensitivities.

In general, select the highest sensitivity variant that provides for the desired full-scale current range. For the design parameters in this example, the TMCS1100A2-Q1 with a sensitivity of 0.1 V/A is the proper selection because the maximum-calculated ±23.2 A linear measurable range is sufficient for the desired ±20-A full-scale current.

After selecting the appropriate sensitivity variant for the application, the zero-current reference voltage defined by the V_REF input pin is defined. Manipulating Equation 28 and using the linear range defined by V_OUT,max, V_OUT,min, and the full-scale input current, I_IN,FS, calculate the maximum and minimum V_REF voltages allowed to remain within the linear measurement range, shown in Equation 33 and Equation 34.

Equation 33.

Equation 34.

Any value of V_REF can be chosen between V_REF,max and V_REF,min to maintain the required linear sensing range. If the allowable V_REF range is not wide enough or does not include a desired V_REF voltage, the analysis must be repeated with a lower sensitivity variant of the TMCS1100-Q1. Equation 28 can be manipulated to solve for the maximum allowable current in either direction by using the selected V_REF voltage and the maximum linear voltage ranges as in Equation 35 and Equation 36.

Equation 35.

Equation 36.

Table 10-6 shows the respective values for the example design parameters in Table 10-4. In this case, a V_REF of 2.5 V has been selected such that the zero current output is half of the nominal power supply. This example V_REF design value provides a linear input current-sensing range of –24.5 A to +22 A, with the positive current defined as current flowing into the IN+ pin.

After selecting a V_REF for the application design, an appropriate source must be defined. Multiple implementations are possible, but could include:

• Resistor divider from the supply voltage
• Resistor divider from an ADC full-scale reference
• Dedicated or preexisting voltage reference IC
• DAC or reference voltage from a system microcontroller

Each of these options has benefits, and the error terms, noise, simplicity, and cost of each implementation must be weighed. In the current design example, any of these options are potentially available as a 2.5-V V_REF is midrail of the power supply, a common IC reference voltage, and might already be available in the system. If the primary consideration for the current application design is to maximize precision while minimizing temperature drift and noise, a dedicated voltage reference must be chosen. For this case, the LM4030C-2.5 can be chosen for to optimize system accuracy without significant cost addition. Figure 10-3 depicts the current-sense system design as discussed.