INA241A: Distortion seen when INA241 powered by 3.3V and sensed input PWM at >8V

Hi,

You’re correct that this is not an expected behavior.

Is the 5V provided by an off-board power supply? If so, can you reduce it to 3.3V and observe the behavior?

In addition to output voltage, please also probe Vs and REF

Regards

Guang

Yes, the Vs is supplied by an off board lab supply. After starting at 5V and reducing down to 3.3V the same behavior is observed. The attached screenshot is at 3.3V Vs. Vs and Vref are stable.

Hi Adam,

What is the PWM frequency? Can you show the input common mode voltage? Maybe zoom in around the peaks to see if there is any pattern.

Can you post the schematic as well? Although I feel something in your setup must changed between the two supply levels. Because if supply is the only variable, we shouldn’t see the issue.

Regards

Guang

PWM frequency is ~300kHz. The input common mode is a square wave alternating between 0V and power supply voltage (In this case typically 12V). You can zoom in on the picture I attached it seems pretty consistent. I don't have a schematic, I am connecting the INA241 dev board input between the output of the filterless class D amp and inductive load. I am using a 33mOhm sense resistor with the 20V/V gain option. I can guarantee you the only variable that is changing is the supply voltage. Even shifting the Vref does not have an effect on distortion.

FYI the Class D amp ic is the MAX9737

Yes, I see that in the datasheet now. Is there any way this sensor would work at 300kHz? Maybe by adding some common mode filtering? What if I power it with 12V (Same as the class D amp)? It's not desirable but would placing an inductor between the sensor input and the class D amp output help? Does TI offer any sensor that is designed to work up to 300kHz?

Also, would going over 125kHz damage the chip or would it just render the "Enhanced PWM rejection" feature unusable?

Another question, would a hall effect sensor be more appropriate for this application?

So if we do not actually need the PWM rejection feature, operating at >125kHz should not be a problem correct? I will look into the hall sensor thanks for the recommendation.