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.

BOOSTXL-DRV8323RS: EVAL-Board DRV8323RS: Power Input wires VM and VDrain blow up.

Part Number: BOOSTXL-DRV8323RS

EVAL-Board DRV8323RS: Power Input wires VM and VDrain blow up.

We use the driver eval board to drive a 7 pole pairs BLDC motor (max power approx. 340W). The commutation is done by a CPU with implemented current controller. We are using the board in 6xPWM mode. All internal control registers of the DRV were left on the default values.

Due to errors in the software, the commutation sequence didn’t start right away after power on, because no interrupt from the HALL sensors occurred. So, the BLDC was standing still and the current gauge showed 0,04 A (40mA nominal driver board current), since there was no updated commutation pattern from the CPU.

Meanwhile the current controller recognized the zero current and tried to reach the desired current by increasing the duty cycle. Soon the duty cycle reached 100%, but was not put through to the driver board due to missing interrupt in the CPU. Then, as I poked the BLDC to trigger an interrupt, suddenly a firework began. The PCB routings V_M and V_DRAIN completely burned away and also the case of the driver shows some burned spots. See the marked positions in figure 01 below.

The source voltage was set to 28V and a current limiter was set to 1,6A and DIDN’T trigger during the whole event (!).

On a second, identical driver board, it accidently happened again. The same conditions. The same spots burned away.

See Figure 1 in the attached file.

Due to project specification, we reworked the board slightly. We removed the three low-sided shunts, bridged the shunt terminals and placed only one shunt on Phase A (0,011 Ohm instead of 3 x 0,007 Ohm) to measure only one current value (see the pictures below). Connection was tested and was also working well before the above-mentioned accident.

Further, the soldering pads beneath the screw output terminals for the phases on the first board were damaged due to vibrations during former testing and needed to be bypassed from the MOTA, MOTB and MOTC test pads (see the pictures below). These connections were checked after soldering too and were also working well before the accident. The second board had no damage on any of the terminals.

We don’t understand what can be the reason blowing up the power supply wires on the DRV8323RS. If there was an extremely high current because of the 100% duty cycle, wouldn’t it blow up the MOSFETs or the shunt in the first place? A high Back-EMF blowing up the DRV because of high voltage peaks on the measure inputs (such as SHA, SHB and SHC inputs) seems also very unlikely to us, since the BLDC didn’t even spin up.

We ask the community about similar experience with the Eval-Board and the component DRV8323RS. Is there any explanation what happened?

Photos.pdf

In the attached file you will find photos of Board 1 and Board 2.

Board 1 is shown in Figure 2, Figure 3 and Figure 4.

Board 2 is shown in Figure 5, Figure 6 and Figure 7.

  • Nobert,

    It's possible that the current that damaged the board came from the on board caps and not the power supply itself. I am obviously not very familiar with your control scheme or wiring so it's hard to comment directly on that. I usually don't advise running higher current tests with this much modification of the EVM.

    Could you check the VM and VDRAIN pins near the device for voltage anomalies when running? Maybe we can identify what is causing the issue.

    VM is a lower current supply used for charge pump and internal logic. VDRAIN is more of a monitor than a supply. I don't see why either of these pins would take higher current unless something else is breaking down inside the chip first, maybe due to the hardware/software modifications made? 

    Regards,

    -Adam

  • Hello Adam,

    sorry for the late answer. I was one week off and after that other project work requests my attention.

    I want to present you our measurements in attached file VM_Measurements.pdf

    VM_Measurements.pdf

    Let us know your opinion!

     Can we change the information flow from forum to email? It would make a lot easier. At the end when we found cause of the problem we can inform the forum community!

    So, now it is the time to come back to the problem. We have to find an answer about the effect we have identified, because we must be sure that the motor drive component is mature and safe. Maybe we had here only a handling problem. Otherwise we cannot use it for flight applications.

    I think I must detail more about the test setup.

    We have a MCU Board (Power PC evaluation board from NXP) which provide us all control signals for the BOOSTXL (BOOSTXL-DRV8323 Eval Board).

    Between the PC Board and the BOOSTXL we have a self-made Adapter Board which contains drivers and glue logic. The BOOSTXL is directly connected on our Adapter Board.

    The following inputs on the BOOSTXL are buffered at the adapter board by a single Schmitt trigger gate from the MCU Board:   ENABLE, CAL

    The following signals are wired directly by the adapter board from the MCU Board to the BOOSTXL: INHA, INHB, INHC, INLA, INLB, INLC

     

    Note also, that the MCU Board and the Adapter Board are powered by an own 12V DC lab power supply.

    The BOOSTXL is powered by an own 28V DC powerful lab power supply. (but current limitation was set.)

    Both lab power supplies are switched on/off manually by the operator.

    The fault appeared when the power for the BOOSTXL was switched off, the Power for the MCU and Adapter Board was switched on. The MCU Board was operating which means:

    a)       Enable was active (high); CAL was low (inactive)

    b)      PWM-Signals are operating on the PWM inputs of the BOOSTXL

    Now the operator has switched on the 28VDC power supply and the disaster happened immediately after switch on.

    Regards

    - Norbert -

  • I forgot to to inform that the PWM on MCU Board was set to 100% PWM.

  • Nobbi,

    There should not be PWM or enable applied to the DRV before the power is applied. This likely caused a large inrush current and damaged the board.

    Regards,

    -Adam