I want to use an AD8418 (http://www.analog.com/media/en/technical-documentation/data-sheets/AD8418.pdf) together with the ADC of the uC to measure the current of a BLDC motor for two phases, that is used for field oriented control. As a driver I use a DRV8313 (http://www.ti.com/lit/ds/symlink/drv8313.pdf) which allows to connect the shunt resistor to the low-side mosfet. My question is now where I should put my shunt resistors to measure the currents:

1.) Directly to the phases of the BLDC motor. This is found in the application note of the AD8418.


2.) Put the shunt resistors to the low side mosfets.

When I put the shunts to the low side mosfets, I need to measure the current only when the low-side mosfet is active. For that, I need center-aligned PWM to trigger the ADC in the right moment. Is this also needed when I put the shunt resistors to the BLDC phases and if yes, when needs the current to be measured? Are there any advantages with method 1.)? The actual PWM signal will be around 25kHz, where the CMRR is about 80dB, that should be still enough.

Here is also a schematic of how I planned to do it with the method 1.)

enter image description here


1 Answer 1


Both methods have advantages and disadvantages. In method 1 you don't need any serious bandwidth, just about 100kHz is enough. Also you get isolated measurement, which is very good. Just remember you can also sample the third phase to detect short circuit. In method 2 you don't get good short circuit protection, your bandwidth should be at least around 2MHz, since current doesn't always appear on shunt resistors, you don't get isolation for free, but you can use cheap and simple components.

So in the end it's your call, depending on application. I used both methods and can't say any of them is always better than the other.

  • \$\begingroup\$ Do you mean bandwidth for sampling with the ADC or PWM for the phase modulation? Do I need center-aligned PWM also with method 1 or is synchronization of the ADC and the PWM not necessary like for method 2? \$\endgroup\$ Commented Feb 12, 2017 at 11:28
  • \$\begingroup\$ I mean on the amplifier before the ADC. ADC itself will have sampling rate same as your PWM, i guess. The center-aligned PWM is not critical for method1, but will probably reduce noises a little bit. \$\endgroup\$
    – user76844
    Commented Feb 12, 2017 at 11:31
  • \$\begingroup\$ By the way, i wouldnt use the sensor on the left for overcurrent. With 25kHz PWM it will see nothing, unless it has 2.5MHz bandwidth. And even then it's not optimal protection... \$\endgroup\$
    – user76844
    Commented Feb 12, 2017 at 11:41
  • \$\begingroup\$ Which sensor do you mean? I actually use field oriented control to adjust the torque and field producing current, so the measurements will only indirectly be used for over current protection. The ADC then needs to measure the current only when used for the FOC loop, so the actual sampling will be around 500Hz. \$\endgroup\$ Commented Feb 12, 2017 at 12:24
  • \$\begingroup\$ I mean the one on duagram you show, on the left. Before the bridge. \$\endgroup\$
    – user76844
    Commented Feb 12, 2017 at 12:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.