This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

DRV8316: Sensored feedback

Part Number: DRV8316
Other Parts Discussed in Thread: DRV8323, MCT8316Z

Hi there, we are working with the DRV8316 and have limited IO on our microcontroller. We need to at least drive the PWM and read the hall sensors

A couple of questions: 

1. Can we run the DRV8316 in 1xPWM mode - like figure 19 on the DRV8323 datasheet illustrates? ie. use the hall sensors from the motors as input to the driver?

2. Do we need all 3 SOX pins to measure current? Can we tie them all together or only use one - we just want to obtain a pseudo accurate motor current reading.

  • Hello Daniel,

    Thanks for the questions and for posting to the MD forum. 

    1) Unfortunately there is not a 1xPWM mode available on this device. It might be worth it to look at MCT8316Z. Which has a built in FET like DRV8316 but offers the built in sensored trapezoidal control that you are looking for. So the input can be driven with either an analog voltage or a PWM input. 

    2) Not all current sense inputs need to be used if you want to get a tototal current measurement and not a measurement per phase you can tie all of them and use one. Make sure to tie unused input current sense pins to ground and float the outputs of the CSA that are not used. MCT8316Z does not have a CSA pin but it is built into the device.

    I hope this helps and let me know if you have any additional questions!

    Best,

    Isaac

  • Thank you!

    Please see my comments below:



    How does the MCT8316Z compare in efficiency to the DRV8316? It seems like it uses a similar power stage.

    Is there a way to get current feedback from the MCT8316Z?



    Can we use the halls as inputs to the DRV8316?

    Can the DRV8316 work in 3xPWM mode with trapezoidal? Is there any drawbacks compared to using 6xPWM?

    By combining the current sense pins on the DRV8316, what would we be reading? The total phase current? Average phase current?

    We are also using the DRV8323 in 6xPWM mode in a senseless application. Is there any drawbacks to running this in 3xPWM?

  • Hey Daniel,

    we are currently out of the office on holiday. We will review your question and submit a response tomorrow. We appreciate your patience!

    Best,

    Isaac

  • Hey Daniel,

    1) The MCT8316Z and the DRV8316 have an identical power stage capable of the 8A peak current. The main difference is the built in control.

    2) All the current sense capabilities are built into the MCT8316Z, and it does not have a pin to sample the current.

    3) Yes hall inputs are  meant to be used via the HPx and HNx pins. Device accomodates for digital output hall sensors as well as differential hall signals.

    4 and 6) Yes DRV8316 does have a 3x PWM option that can work to create a trapezoidal algorithm. The main benefits of using 3x PWM vs 6x PWM is reducing the amount of connections needed from MCU to the DRV device and it can also simplify your algorithm since you have a lesser amount of signals to worry about. The main downside of using 3x PWM mode is that you are limited by the devices logic table in how the remaining gates are controlled, so you do not have full control of each MOSFET in this kind of topology. This can be a downside if your algorithm needs full control of each and every single gate.

    5) Just need to correct my previous comment. I forgot that the conversation was on DRV8316, so there is no way to ground the current sense inputs since they are internal to the device and  cannot be tied together before they are input into the device. SOx outputs should not be tied together, since this can yield unpredictable SOx voltages. I just wanted to clear up that statement for you since I believe I may not have been clear enough. You are still however allowed to float the current sense outputs if you are not using them but you would only be able to obtain current from one phase of the motor. If you are able to free up one ADC output you can sample 2 phases by using Kirchhoff’s Current Law Ia + Ib + Ic = 0.

    Best,

    Isaac

  • 1. How does the MCT8316Z compare in efficiency to the DRV8316? It seems like it uses a similar power stage.

    The MCT8316Z and the DRV8316 have an identical power stage capable of the 8A peak current. The main difference is the built in control.

    Understood

    2. Is there a way to get current feedback from the MCT8316Z?

    All the current sense capabilities are built into the MCT8316Z, and it does not have a pin to sample the current.

    Understood. Our application requirements require us to know the motor current to detect open and stall events so this part my not be usable.



    3. Can we use the halls as inputs to the DRV8316?

    Yes hall inputs are  meant to be used via the HPx and HNx pins. Device accomodates for digital output hall sensors as well as differential hall signals.

    I was talking about the DRV8316 and not the MCT8316Z. Please address. 

    4. Can the DRV8316 work in 3xPWM mode with trapezoidal? Is there any drawbacks compared to using 6xPWM?

    Yes DRV8316 does have a 3x PWM option that can work to create a trapezoidal algorithm. The main benefits of using 3x PWM vs 6x PWM is reducing the amount of connections needed from MCU to the DRV device and it can also simplify your algorithm since you have a lesser amount of signals to worry about. The main downside of using 3x PWM mode is that you are limited by the devices logic table in how the remaining gates are controlled, so you do not have full control of each MOSFET in this kind of topology. This can be a downside if your algorithm needs full control of each and every single gate.

    Understood. Does this have any impacts on efficiency? 

    5. By combining the current sense pins on the DRV8316, what would we be reading? The total phase current? Average phase current?

    Just need to correct my previous comment. I forgot that the conversation was on DRV8316, so there is no way to ground the current sense inputs since they are internal to the device and  cannot be tied together before they are input into the device. SOx outputs should not be tied together, since this can yield unpredictable SOx voltages. I just wanted to clear up that statement for you since I believe I may not have been clear enough. You are still however allowed to float the current sense outputs if you are not using them but you would only be able to obtain current from one phase of the motor. If you are able to free up one ADC output you can sample 2 phases by using Kirchhoff’s Current Law Ia + Ib + Ic = 0.

    Understood. So using the DRV8316, we cannot short pins together. Does using two phases result in additional accuracy compared to 1 phase? We only want to use this current sense for average motor current telemetry - detecting open, stalls, and using it for thermals.

    6. We are also using the DRV8323 in 6xPWM mode in a senseless application. Is there any drawbacks to running this in 3xPWM?

    No answer provided.

    Let me know

  • Hey Daniel,

    Here is the response to the remaining questions:

    3) Can we use the halls as inputs to the DRV8316?

    Unfortunately DRV8316 does not have input for hall sensors. For a sensored application, the hall sensors would have to go through MCU GPIO to check the states of the motor then MCU PWM outputs to the DRV8316 INHx and INLx inputs. 

    4) Does this have any impacts on efficiency? 

    Efficiency depends on the algorithm that you are running. 3x PWM still allows you to run an FOC algorithm which is considered to be the most efficient but can also run trapezoidal control. As far as if a  6x PWM FOC algorithm being more efficient than a 3x PWM FOC algorithm, I have not done enough testing to see what the differences in efficiency would be, but unfortunately this can vary through algorithm.

    5) Does using two phases result in additional accuracy compared to 1 phase? We only want to use this current sense for average motor current telemetry - detecting open, stalls, and using it for thermals.

    You can obtain average current from reading the current of the phases. If you are using a trapezoidal commutation method then current is present in two phases and the third one will not have current, this is due to the nature of the trapezoidal algorithm where only to phases are being driven at a time. Typically having it two phases can give you a more accurate reading of what is occurring in the motor. But if you are not using the current to commutate the motor and just monitor the current then it should be possible to just use one CSA.

    Here is a Texas Instruments Precision Labs video that might be helpful: https://training.ti.com/ti-precision-labs-motor-drivers-peak-vs-rms-current

    6) We are also using the DRV8323 in 6xPWM mode in a senseless application. Is there any drawbacks to running this in 3xPWM?

    The benefits and drawback are the same as listed in 6x PWM vs 3x PWM for DRV8316. This will reduce the amount of PWM signals needed but if you are using a very specific commutation algorithm that does not coincide with the 3x PWM logic table presented in the datasheet, then you might need to revaluate performance with a completely new commutation method. As far as efficiency, this is debatable depending on the algorithm being used.

    Best,

    Isaac

  • Hi there,

    To clarify: there is not a clear difference between 3x trapezoidal and 6x trapezoidal efficiency correct?

    Can you use the DRV8316 in a sensorless trapezoidal application in 3xPWM mode?

    Can we place a small current shunt (around 4 mOhms) in between PGND of a MCF8316/MCT8316 and our system GND to measure the overall current running through the motor and driver? Is there any problems with this?

  • Hi Daniel,

    To clarify: there is not a clear difference between 3x trapezoidal and 6x trapezoidal efficiency correct?

    Correct there is not a clear difference between 3x trapezoidal and 6x trapezoidal efficiency.

    Can you use the DRV8316 in a sensorless trapezoidal application in 3xPWM mode?

    Yes, DRV8316 can be used in a sensorless trapezoidal application in 3x PWM mode.

    Can we place a small current shunt (around 4 mOhms) in between PGND of a MCF8316/MCT8316 and our system GND to measure the overall current running through the motor and driver? Is there any problems with this?

    It is possible to do this there are just some things that need to be taken into consideration. Just a couple of notes of things to consider:

    Please ensure that the shunt resistor used to measure the current is a very low value and has very low inductance. This is because the difference between PGND and GND cannot be more than 0.3V or else device will not operate. Therefore, shunt voltage ( Current x 4mOhms ) needs to be less than 0.3V. This should not be hard to achieve since MCx8316 devices are rated for 8A peak. Also ensure negative transients on OUTx do not exceed 0.7V or else the device will not work.

    I hope this helps!

    Best,

    Isaac