• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 31 subscribers
  • Views 363 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

DRV8231A: DRV8231A Failure and Operational Conditions

Conor
Guru 12115 points
Part Number: DRV8231A

Tool/software:

Hi,

We are currently evaluating the DRV8231A for replacing a discrete circuit.
However, during operation, I have encountered an issue where the IC fails under certain conditions. Below is a summary of the application and the problem:

Application Details

  • Motor specifications: Brushed DC motor, 24V, 0.2A (inrush current: ~2A).
  • Control pattern: Forward and reverse rotation every 2 seconds.
  • VM power supply and control signals: 2ms after VM is input, the input control signal and VREF (5V) rise with a delay.

Problem Description

  • After a certain period of operation, the motor unexpectedly stops, and upon inspection, I found that the resistance between VM and GND on the IC measures approximately 120Ω.
  • This behavior suggests a possible internal failure of the IC.

Questions

  1. Regarding VM power and control signals:

    • 2ms after VM is input, the input control signal and VREF (5V) rise with a delay. Could this affect the stability or operation of the DRV8231A?
    • Are there any recommended timing or voltage conditions for VM and control signal initialization to ensure reliable operation?

  2. Regarding potential cross-conduction (shoot-through) and internal protection features:

    • I suspect that cross-conduction might be the cause of the IC failure. Does the DRV8231A have built-in protection to prevent shoot-through?
    • If such a protection mechanism exists, under what conditions might it fail to prevent damage to the IC?

  3. Regarding grounding (body earth):

    • In this prototype, the body ground is not connected. While I understand that this primarily affects thermal dissipation and noise immunity, could the absence of a proper body ground contribute to this type of failure?

I would appreciate your insights into these questions to help identify the root cause of this issue and implement necessary design improvements.

Thanks,

Conor

  • 0
    TI__Genius 11711 points

    Hi Conor,

    Thanks for your question.

    I have assigned this request to the device expert. He will review the question and reach out to you.

    Thanks,

    Ibinu 

  • 0
    TI__Mastermind 19380 points

    Hi Conor,

    The VM and control signal timings seem to be within spec and should be fine, although it seems the inputs and VREF are exceeding recommended operating conditions they are still within abs max specs and should be fine. 

    All TI motor drivers have internal shoot-through protection. This device has a 200ns dead time, where both high-side and low-side FETs are off, to prevent a shoot-through condition. 

    It's not typical to drive a motor between the forward and reverse state so often. If using PWM inputs, its recommended to switch between a forward and brake or forward and coast state. 

    When switching to a brake or coast state, this allows the motor current a safe path to decay from the load during PWM off time. It's likely by switching repeatedly from forward to reverse the motor current builds and doesn't have a path to decay. 

    When you say, "the body ground is not connected" do you mean the GND pin (pin 7) is not connected, or the thermal pad is not soldered to PCB? 

    Best,

    David

  • 0 in reply to David Medis
    Guru 12115 points

    Hi David,

    Thank you for your reply.

    David Medis said:
    When you say, "the body ground is not connected" do you mean the GND pin (pin 7) is not connected, or the thermal pad is not soldered to PCB? 

    By body ground I meant the thermal pad. Is it possible that not connecting the thermal pad is related to this failure? Also, if you have any other ideas or things to check, I would appreciate your advice.

    Thanks,

    Conor

  • 0 in reply to Conor
    TI__Mastermind 19380 points

    Hi Conor,

    "The device thermal pad should be attached to the PCB top layer ground plane and internal ground plane (when available) through thermal vias to maximize the PCB heat sinking." - From section 11.1 Layout Guidelines of datasheet. The thermal pad connection is used for heat dissipation. As long as the GND pin of the device is connected to system ground of PCB there should be no grounding issues. 

    As stated in previous post, it's not recommended to switch between the forward and reverse states for long periods of time without having a safe path for current to decay during PWM off times. 

    What's the application where this type of operation is needed?

    If possible, please provide more detail into how many units are seeing this issue? 

    Conor said:
    After a certain period of operation, the motor unexpectedly stops

    What is the operation and how long is it? Is there any load on the motor? 

    Best,

    David

  • 0 in reply to David Medis
    Guru 12115 points

    Hi David,

    Thank you for your great feedback!

    David Medis said:
    The VM and control signal timings seem to be within spec and should be fine, although it seems the inputs and VREF are exceeding recommended operating conditions they are still within abs max specs and should be fine. 

    You state that "the input signal and VREF exceed the recommended operating conditions", but the input signal is 5V and VREF is 5V.

    From the data sheet, the input signal is within 5.5V, so there doesn't seem to be a problem. Does VREF need to be within 3.6V? The pin description in the data sheet seems to indicate that it supports up to 5V so I'm a little confused as to which is correct.

    David Medis said:
    It's not typical to drive a motor between the forward and reverse state so often. If using PWM inputs, its recommended to switch between a forward and brake or forward and coast state. 

    Currently, IN1 and IN2 are used and the control is set to complete with the brake. However, the brake and VM cut-off are closely related, so it is possible that the attenuation path you mentioned may disappear.
    Please let us know if there are any regulations regarding the brake hold time.

    David Medis said:
    When switching to a brake or coast state, this allows the motor current a safe path to decay from the load during PWM off time. It's likely by switching repeatedly from forward to reverse the motor current builds and doesn't have a path to decay. 

    The defective product is not the cross-connected FET, but the low resistance between the upper and lower FETs has caused the damage.

    Which FET does tDEAD turn off? Also, it would be helpful if you could comment on the possibility of the upper and lower FETs having low resistance.

    David Medis said:
    What is the operation and how long is it? Is there any load on the motor? 

    The application is E-Lock. We are considering replacing the E-Lock control from a discrete circuit to DRV8231A. The control target is a brushed DC motor, about 24V/0.2A (inrush 2A). During the prototype stage, a failure occurred during a test that repeatedly rotated forward and backward at 2-second intervals.

    Thanks,

    Conor

  • 0 in reply to Conor
    TI__Mastermind 19380 points

    Hi Conor,

    Yes, the input voltages should not cause any issue. They are higher than the "recommended operating conditions" but still under the abs max conditions so they should be fine. 

     I believe I understand now. The inputs are IN1 = 1 and IN2 = 0 (forward) for 2 seconds and switching IN2 = 1 (brake) before going in the reverse direction, is this correct? If so, this should not be an issue and this driver is capable of driving under those conditions. Please see this app note for more information on current recirculation techniques.

    Under what conditions is the resistance between VM and GND being measured? Is the device on, off, outputs in forward, reverse, hi-z, etc.?

    How many devices are seeing an issue and how many have you tested? Are these tests being conducted on EVM or personal design?

    During tDEAD both high-side and low-side FETs are OFF to prevent shoot-though. 

    Best,

    David