• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 780 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

DRV8873H-Q1EVM: DRV8873H-Q1 output instability

TISL
Expert 3795 points
Part Number: DRV8873H-Q1EVM

Dear TI team,

See below my customer issue with our motor driver EVM:

__________________________________________________________________________________________

I noticed a problem with the DC motor driver DRV8873H-Q1. Below is the context and the observations done.

 

Context:

- EVM board DRV8873H-Q1EVM (hardware version, board modified to set all slew rates availables with an external resistor, and resistors R30/R32 removed to control the driver with an external PWM done by a function generator)

- 18V power supply

- DC motor: R=3.9ohms, L=5mH (blocked for the demonstration)

- F_PWM=40kHz

- Mode PH/EN (set on GUI software): PH=0, EN=0/1

- Slew rate set to 13V/µs

- Motor connected between OUT1 and OUT2

- Wake: ENABLE

- Output states: ENABLE

- Open load detection: ENABLE

- Itrip control: ENABLE (Itrip=6.5A, Toff=20µs)

 

Observations:

Duty cycle = 22%:

By applying external PWM (blue curve) the motor current (green) increases. The OUT1 (yellow) falling slew rate is around 13V/µs whereas the rising slew rate is much higher.

 

Duty cycle = 24%:

When the OUT1 seems reach 0V, the driver becomes unstable. The rising slew becomes unstable, the motor current starts to oscillates and the motor starts to vibrate. A video linked was done to show this phenomenon.

 

Beyond this specific duty cycle, the driver and motor behave well.

This phenomenon was observed first by driving the motor with the GUI software and was reproduced with external PWM control.

This phenomenon is also present in PWM mode by alternating the outputs between H/L and H/H states (does not append in PWM mode by alternating the outputs between H/L and Hi-Z/Hi-Z states).

This phenomenon is much more present with low slew rates configurations and vibrations of the motor are stronger.

This phenomenon was reproduced with a power resistor instead of a motor on the outputs.

 

Did you already noticed this behavior or could it be a default of my EVM?

Where does it come from?

Does it happen with the DRV8873S driver too?

 

Thanks you for your interest

_________________________________________________________________________________________

  • 0
    TI__Guru** 114905 points
    Hi TISL,

    We will investigate and should have an initial response by Monday.
  • 0 in reply to Rick Duncan
    TI__Guru** 114905 points
    Hi TISL,

    The motor acts like it is trying to start but does not have enough current at 24%.

    If the customer tries to operate the motor with a DC supply, does the motor behave similarly when ramping the supply between 2 and 6V?

    Would you please ask the customer to capture several individual scope shots at 24% without persistence?
    Also please ask for a zoomed out scope capture (maybe 500us/div). We would like to see if there is a repetitive pattern in the current.
  • 0 in reply to Rick Duncan
    Prodigy 40 points

    Hello Rick, thank for the reactivity, I am the costumer who noticed this issue.

    Below are my answers.

    If the customer tries to operate the motor with a DC supply, does the motor behave similarly when ramping the supply between 2 and 6V?

    -> No, when ramping up the voltage, the motor behaves normally, there are no vibrations and the current is stable (motor blocked).

    Would you please ask the customer to capture several individual scope shots at 24% without persistence?

    Also please ask for a zoomed out scope capture (maybe 500us/div). We would like to see if there is a repetitive pattern in the current.

    -> Below are some oscilloscope shots with the same setup at 24%:

    As before, Blue=command, Yellow=OUT1 (OUT2=high), Green=motor current.

    Below are 2 single shots (without persistence) at 2ms scale:

    Below are two single shots (without persistence) at 400µs scale:

    Below is a single shot (without persistence) at 100µs scale:

    On the image above at 100µs/div, I looked the rise time on the output transitions. The pattern alternate between fast rise time (1) and slow rise time (0) with the following sequence:

    001111011101000001111111010101000001101

    Below are zoomed image on fast and slow rise times:

    Hoping that these pictures are clear sufficient. Thank you for your time.

    Alexis Monaci - Electronic engineer

  • 0 in reply to Alexis Monaci
    TI__Guru** 114905 points
    Hi Alexis,

    The images show that the slew rate and propagation delay appear to change. Both are based on the SR pin.

    Can you try setting the SR to a different value a non Hi-Z value (68k or 22k to GND)?
  • 0 in reply to Rick Duncan
    Prodigy 40 points

    Hi Rick,

    I tried with the 68k config (SR=18,3V/µs), the behavior seems to be the same but less strong, see below:

    400µs scale:

    The rising slew rate alternate between fast and slow:

    With the 22k config (34V/µs) the phenomenon is not visible at Vm=18V. However if I increase the voltage to 29,4V (in our application we use a battery pack from 18V to 29,4V; I only mention 18V to stay simple and focus on the driver itself), the phenomenon appears:

    100µs scale:

    The rising slew rate alternate between fast and slow:

    With the GND config (53,2V/µs) the phenomenon is not visible at Vm=18V nor 29,4V.

    Thank you for your support!

    Alexis Monaci - Electronic engineer

  • 0 in reply to Alexis Monaci
    TI__Guru** 114905 points
    Hi Alexis,

    So this appears to be some type of noise affecting the slew rate. Let me check with the design team to determine if there is a recommendation to reduce the noise. I hope to have an answer by Friday.
  • 0 in reply to Rick Duncan
    TI__Guru** 114905 points
    Hi Alexis,

    After speaking to the design team, it was suggested to try placing a 1nF capacitor on the SR pin as a filter.

    This may be necessary when operating at 7.9, 13, 18.3, or 34 V/us slew rates. It may also depend on layout.