DRV591: selection of inductance in DRV591

Hello Lilian! What is the intent of reducing the inductance in this application and what frequency are they switching at? The performance of the device as shown in the datasheet is based on the reference schematic design. Switching frequency and the clause of the LC filter design will impact performance. Please reference the page 11 OUTPUT FILTER CONSIDERATIONS for a good description of this relationship.

Shawn can then help you with any further questions. Thanks, Jeff

Hi Lilian,

LC filter on the output is to supress the ripple on the output current and EMI. So if lower value inductors are used for it, we need to evaluate it from these two perspectives.

The first, you need to know how much ripple current can be acceptable on the output. Higher ripple current reduces the actual temperature differential on TEC. On page 11 and 12 in the DRV591 datasheet, you could find how to calculate the ripple current from both frequency domain and time domain. For this calculation, you need to know how much VDD, TEC resistance and maximum ouput current are. Please calculate for the worse case(50% duty-cycle), and reserve enough margin for the design. Basically, the ripple current increases 4.55 times with changing 10uH inductors to 2.2uH in the same application.

To achieve a better EMI performance, please place the LC filter close to the output of DRV591 as much as possible. And it's very hard to evaluate the EMI performance degradation in a system, usually it's done by performing experiments. Because EMI performance is a system leve issue, and the result could be very different between the different systems. If your system is sensitive to EMI, please just keep 10uH inductors in your design.

Best regards,

Shawn Zheng