DRV10987: What TI parts can you recommend to drive 24-32V motor, at 40A (pp)? using sinusoidal or Trapezoidal control.

Hi Bhaktprakash,

Thank you for your post!

1. Here are a few resources to help understand the differences between sinusoidal control and trapezoidal control:
   1. https://training.ti.com/ti-precision-labs-motor-drivers-comparison-commutation-methods?context=1139747-1138777-1139742-1140395
   2. TI Precision Labs - Motor Drivers: Sinusoidal Control | TI.com Video
   3. TI Precision Labs - Motor Drivers: Trapezoidal Commutation | TI.com Video
   4. [FAQ] Trapezoidal Motors vs. Sinusoidal Motors - Motor drivers forum - Motor drivers - TI E2E support forums

2. For your design do you need to use integrated control? Or would you consider using external control? For integrated control, unfortunately the only device that supports sinusoidal commutation over 24V we no longer recommend for new designs. Additionally, for your current requirement of 40A you won’t be able to use a device with integrated FETs since the current limit for those devices are too low for your design. One solution is to use a gate driver that supports external control with external MOSFETs. This would allow you to use a sinusoidal algorithm in the MCU to control the driver. If you need integrated control for your design then the option I would recommend would be to use a trapezoidal control gate driver with external MOSFETs.

3. For your design, are you using sensored or sensorless control?

The answer to these questions will help me understand which gate drivers might be best for your situation.

Regards,

Anthony

Hello Anthony,

Thank you for your response.  I am a colleague of Bhakt's working on the same project.  I'm hoping I can answer your questions and provide a little more information regarding our application.

2.) External control will work for our application.  We need a driver that is suitable to work with external FETs.

3.) We currently have feedback from hall sensors and may incorporate an encoder going forward.

Additional information regarding our application.  We are looking to set a target current, maximum 60V peak to peak (30A at a time through any one FET), and have the driver drive our motor at that current.  We have previously seen this defined as direct torque/current control in several applications, but have not located a TI part that uses this terminology. 

I will admit, I am not entirely clear on the correlation between driving a motor this way and the commutation method used.  Does trapezoidal or sinusoidal have an advantage when trying to achieve this?  Our basic requirement is to be able to achieve very large acceleration with our BLDC motor in a short amount of time (<10's of milliseconds).

We have placed on order the DRV8300DRGE-EVM development board, see link below, to hopefully get us started.  

https://www.digikey.com/en/products/detail/texas-instruments/DRV8300DRGE-EVM/13692792?s=N4IgTCBcDaKHAEYCcA2AtAEQEoDUAcBmABkKwHEBRVc7AWVQDl04QBdAXyA

Looking into the documentation, however, I'm still uncertain we will be able to use this dev kit to drive our motor with the direct torque/current control scheme described above.  

Based on this information, do you have any specific drivers/development kits you could recommend for our application? 

Best,

Noah

Hi Noah,

Thank you for the additional information! I have a few more questions to help me get a little better idea on what would be best for your application:

1. If the motor is a 24V motor, what is your plan in not exceeding the abs max ratings of the motor when you supply 60V peak to peak? 

2. What is the voltage of the power supply that you plan on using? 

Regarding control method, from an efficiency perspective, the most efficient method of commutation between trap and sinusoidal is based on the shape of the back emf of the motor. If the back emf is sinusoidal in shape, then a sinusoidal algorithm would be the most efficient algorithm to use. If the back emf of the motor is more trapezoidal in shape then the most efficient algorithm to use would be a trapezoidal algorithm.

From a torque perspective, trapezoidal control is better at applying more torque to the motor compared to sinusoidal, since the sinusoidal algorithm is not as optimized for torque and is more focused on speed control. If you wanted to maximize the torque you could consider using a Field Oriented Control (FOC) method, but that is a much more complicated method of commutation compared to trapezoidal commutation or sinusoidal commutation. 

For more information on the advantages and disadvantages of the different commutation methods, see this video here: TI Precision Labs - Motor Drivers: Comparison of Commutation Methods | TI.com Video

the DRV8300EVM used in conjunction with the LAUNCHXL-F280049C seems like it would be a decent device to test your motor on, however the EVM is designed for a max current of about 25A, so you wouldn't be able to drive your motor at full speed using that particular EVM. The commutation method for that EVM is sensored trapezoidal control, so it would allow you to test out the performance of your motor for currents under 25A using trapezoidal control. 

Regards,

Anthony 

Hello Anthony,

Thank you, this is very helpful.  I'd like to discuss our design in more depth to help identify an optimal solution, but don't want to get into any more detail on the forum.  Is it possible to move to a private chat or email for further discussion?

Thank you,

Noah

Hi Noah,

Definitely! I just sent you a message with my contact info. 

Regards,

Anthony