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.

DRV8343-Q1: The main reason for the damage of smart driver, such as drv8343s

Part Number: DRV8343-Q1

Hi,we recently use drv8343s for PMSM sensorless control of a 24V system.

However, the driver in the test is easy to be destroyed.

We use the infineon Mosfet of IPI120N06S4-02, whoes Cgd is typically 15nC.

We start with 200mA IDRIVEP, and 3us TDRIVE.

I find the limit value for SHX and DLX as below

Does it mean that if I choose a big IDRIVE, it will easliy cause a big dv/dt, which will cause high spike at SHX and DLX?

It seems that there is not that much region for  SHX and DLX .Some other drivers which use external bootstrap has a big limit for SHX and DLX,like 60V.

Is the voltage spike at SHX and DLX the most common reason for the damage of smart driver?

So we must slowly turn on the MOS to make sure the spike at SHX and DLX not more than Vm+7V? It will increase the power loss of MOS a lot.

Regards

Arrow

  • Hello Arrow,

    Choosing the right IDRIVE setting is crucial to ensure that an appropriate gate drive current is used to turn on the FETs accordingly without introducing dV/dt coupling, ringing, or shoot through conditions from overstressing or overdriving the FETs. 

    From a high level, a big IDRIVE will increase the slew rate and reduce switching losses, but increase the EMI and run the risk of dV/dt coupling, ringing, or shoot-through conditions. A lower IDRIVE will decrease the slew rate and increase switching losses, but decrease the EMI and any possibilities of unintended behaviors from inductance in the powerstage. 

    TI recommends calculating the IDRIVE according to this FAQ information here. For Qgd = 15nC, and a recommended "fast" rise time of 100ns, we would recommend starting at an IDRIVE setting of 15nC/100ns = 150mA. Try starting at the next smallest setting (IDRIVEP_Lx/IDRIVEP_Hx = 65mA = 0111b), then going to higher settings from there. 

    Answers to questions below:

    It seems that there is not that much region for  SHX and DLX .Some other drivers which use external bootstrap has a big limit for SHX and DLX,like 60V.

    This part is rated for VM = 60-V abs max, but we recommend this part mostly for 12-V or 24-V automotive applications, simply because we recommend BLDC gate drivers that are rated for 1.5x to 2x their nominal motor voltage rating to have headroom for any transients or supply increases based on motor and supply inductance. Voltage limits for SHx and DLx are for 60V max since that is the max VM this device can provide. 

    Is the voltage spike at SHX and DLX the most common reason for the damage of smart driver?

    For the DRV8343-Q1, transients that are greater than VCP volts at SHx and DLx are the most common reason for damage around the VM/VCP pins due to a sneak path for the VM pin to have a large inrush of current. This is covered in "Generator Mode" section 10 of the datasheet. 

    So we must slowly turn on the MOS to make sure the spike at SHX and DLX not more than Vm+7V? It will increase the power loss of MOS a lot.

    It is okay to start with smaller IDRIVE settings for now to see how the gate drive affects the performance of SHx/DLx. What voltage are you using for the motor? If you are using your own board as well, there are a lot of decisions you can make to reduce the parasitic inductance of the powerstage to reduce positive/negative transients at the FETs when switching. 

    Thanks,
    Aaron 

  • Hi Aaron

    Thanks for your clarification.

    There are still some confusion for me.

    This part is rated for VM = 60-V abs max, but we recommend this part mostly for 12-V or 24-V automotive applications, simply because we recommend BLDC gate drivers that are rated for 1.5x to 2x their nominal motor voltage rating to have headroom for any transients or supply increases based on motor and supply inductance. Voltage limits for SHx and DLx are for 60V max since that is the max VM this device can provide. 

    Accroding to the 7.1 Absolute Maximum Ratings, the max SHX is relataive to VM +7V, if VM is 60V ,than SHX can go to 67V for Transient 200-ns. if VM is 24V, than SHX can goto 31V. Am I right?

    For the DRV8343-Q1, transients that are greater than VCP volts at SHx and DLx are the most common reason for damage around the VM/VCP pins due to a sneak path for the VM pin to have a large inrush of current. This is covered in "Generator Mode" section 10 of the datasheet. 

    In section 10,"Generator Mode" can cause the ESD D4 break when SHX goes more than VCP, and VCP -VM more than the rated voltage of ESD didoe. But this situation is for which a switch between VM and VDRAIN in figure 67.If I connected VM and VDRAIN together, than the VCP -VM will never exceed the ESD rated voltage.Am I right?

    It is okay to start with smaller IDRIVE settings for now to see how the gate drive affects the performance of SHx/DLx. What voltage are you using for the motor? If you are using your own board as well, there are a lot of decisions you can make to reduce the parasitic inductance of the powerstage to reduce positive/negative transients at the FETs when switching. 

    It is a 24V system with rated power of 1500W, and I'm using my own board. Yes,indeed there are a lot for me to reduce the parasitic inductance. I'm working on it.

    Regards

    Arrow

  • Hi Arrow,

    I double checked and you are right, VM + 7V is with respect to VM because there are internal diodes in the device to ensure that the voltages of the diodes do not exceed 7V. So for the 24V application, this would be 31V.

    The Generator Mode diagram described that if VCP-VM > the abs max voltage of the device, then the ESD diode D4 will conduct, regardless if there is a switch or not. There will need to be a large external load on the motor (generator mode) or large transient to cause SHx to conduct the ESD diode D4. 

    Typically a well-designed powerstage and controlled testing will cause the issues above to not occur. Please let us know if you are in need of any resources in your design and bringup. 

    Thanks,
    Aaron

  • Hi Aaron

    1.  Thanks for your clarification of first question.

    2.  About generator mode or large transient on SHX, if we connect the VM and Vdrain together, the transient on SHX will go through the upper MOS's freewheel diode to VM? Then the VM will be clamped to the V_SHX - V_diode,and VCP-VM will never exceed the limit. Am I right?

    3. Any suggestion for us If we want set the middle IDRVIE, for example between 65mA(0111b) -200mA(1000b), like 120mA?Can we use the external Rgate?If we can, is there a rough formula to calculate the Rg to limit the IDRIVE,  based on the para. of MOS?

    Regards

    Arrow 

  • Hello Arrow,

    The team is on holiday due to Memorial Day in the US, and will get back to you tomorrow!

    Thanks,

    Matt

  • Hi Arrow,

    2. VDRAIN is the sense pin for the high-side MOSFET. The concern about generator mode is the ESD diode (D4), if VCP-VM exceeds the limit then this diode will blow up and results in damage to the device. The current path through the body diode of the HS MOSFET on the powerstage is outside of the gate driver and as long as enough bulk capacitance is placed in the system to absorb transients from entering into the supply voltage VM, then there is no concern about VCP-VM exceeding the limit. So you are right in that analysis. 

    3. You can use a middle IDRIVE as described. Many customers go right to the max IDRIVE setting and potentially damage the driver or FETs due to having too high of a gate drive current, so starting lower is always preferred to ensure longevity of the components. 

    Gate resistance is tricky to calculate since it is a complex relationship between inductance, gate drive current, and gate capacitance. You can use this app note to help: https://www.ti.com/lit/an/slla385a/slla385a.pdf?ts=1622554565795&ref_url=https%253A%252F%252Fwww.google.com%252F

    We always recommend a 0-ohm resistor to begin with, and then increment maybe 5 ohms or so if less gate drive current is needed. We recommend adding gate resistance if you need a gate drive current lower than the smallest IDRIVE setting, otherwise lowering the IDRIVE value should suffice. 

    Thanks,
    Aaron

  • Hi Aaron

    Thanks for your suggestion.

    Sorry for futher question as below:

    You can use a middle IDRIVE as described. Many customers go right to the max IDRIVE setting and potentially damage the driver or FETs due to having too high of a gate drive current, so starting lower is always preferred to ensure longevity of the components. 

    Usually, if the driver is damaged, there must be some reason, such as some para. exceeds the absolute max rating value.

    Theoretically, for example, for a high drive current, what exceeding para. possibly will end up with for the driver?

    Regards

    Arrow

  • Hi Arrow,

    In my experience, high gate dive current results in a fast rise & fall time on the external MOSFETs. This causes ringing and oscillation on the gate drive outputs which exceeds the abs max ratings of the driver.

    Thanks,

    Matt