• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 63 subscribers
  • Views 1293 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.

TPS28225: TPS28225 failure

Wim2532
Intellectual 330 points
Other Parts Discussed in Thread: TPS28225

r: TPS28225

Dear, 

Attached you will find a schematic of a TPS28225 based design, used as a constant current source. Vin 25..26Vdc, Vout 0..24dc, Iout 0..10A. Controlled by using a 226kHz PWM signal. Feedback to a microcontroller by voltage and current measurement. 

The schematic shows a 5V supply voltage to the TPS28225, in reality this is 4.8V. This voltage is not ideal but it works at the customer.

The design works well, having a 6.8 Ohm load connected to the output. By using a 226kHz PWM signal the output can be set between 0 to 24V, with a current between 0..3.5A. When doubling the output current (so the load is 3.4 Ohm) the TPS28225 starts to smoke at 10% duty cycle (output voltage is around 3V than). So the TPS28225 gets burned. 

Normally I would expect the T7 and T8 to be burned when the current gets too high,but not gate driver. 

Questions: Can you explain this? Could the be related to the bootstrap capacitor / circuit. See a Fairchild application note as well (attached in the zip file). Or what else could cause this problem?

Are there any good other reference to build a constant current source, based on this device?

Thanks in advance and kind regards,

Wim van der Steeg 

  • 0
    TI__Mastermind 23155 points

    Hello Wim,

    Thank you for your interest in the TPS28225 driver for your application. I am an applications engineer at TI and will work to resolve your concerns.

    To be able to address the cause of the driver overheating in certain load conditions, it will be important to review the scope waveforms of the circuit in normal operating conditions (no issue with operation) and conditions where there is the overheat concern.

    Without seeing the waveforms, it will be difficult to assume what may be causing the increased power dissipation in the driver.

    Can you confirm the following? VDD at the driver during normal operation, and conditions of overheat.

    Scope waveforms of Phase, Ugate, and Lgate during normal operation and condition of overheat. These waveforms can be taken with scope probes all referenced to ground.  Take waveforms with enough detail to examine if there is deadtime between the high side and low side. Confirm that the frequency and duty cycle is stable. If the frequency or duty cycle is not stable record waveforms with enough switching cycles to show pulse width or frequency instability.

    Based on the CSD18503, the boot capacitor value does appear to be adequate; but the scope waveforms will show if the UGATE level has excessive ripple.

    For the question on reference designs for current source, I will have to review the TI designs to see if there is a current source reference design.

    Regards

    Richard Herring

  • 0 in reply to Richard Herring
    Intellectual 330 points

    Dear,

    We are in the process of capturing the signals as mentioned in the first answer. It would be interesting to know how these signals would look like in a overhead condition and what counter measures to take. Is there some advanced information about this? 

    Thanks and regards,

    Wim van der Steeg

  • 0 in reply to Wim2532
    TI__Mastermind 23155 points

    Hello Wim,

    For the signals plots requested, and how they may look in an overheat condition. I would like to see the plots to confirm if there is any gate drive signals or waveforms on the switch node that may lead to excessive loss. Such as possible cross conduction of the power train, false triggering of the driver from noise which may result in unexpected high frequency switching, etc.

    Do you have any waveforms of the issue, or have you resolved the concern?

    Regards

    Richard Herring

  • 0 in reply to Richard Herring
    Intellectual 330 points
    Hi Richard,

    Last Friday I resolved the issue. The gate series resistor value is too high (47Ohm) and this causes slow switching behavior of the mosfets. Lowering the value to 10Ohm solves the issue. Furthermore, the whole design in on the edge of the limits. So I advised to do a complete redesign.

    Regards,

    Wim
  • 0 in reply to Wim2532
    TI__Mastermind 23155 points
    Hello Wim,
    Thank you for the confirmation. With the TPS28225 adaptive shoot through protection, high value gate resistors can result in some VGS overlap between the high side and low side power devices.
    Can you confirm that the thread is closed so this information is reflected as resolved for the other users on E2E?
    Thank You
    Richard Herring
  • 0 in reply to Richard Herring
    Intellectual 330 points
    Richard,
    Thanks for confirming my experience. It resolves the issue. Thread can be closed.
    Regards,
    Wim