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.

DRV8837: PSPICE model - Coasting and brake current simulation

Part Number: DRV8837

Hi

I am working on a large volume project for a client. Due to the FDA approval I need to go into details with the H-Bridge.
I have downloaded the newest (and only) PSpice model DRV8837 PSpice Model (Rev. A) "SLVM958A.ZIP (18 KB) - PSpice Model"
Analyzing the sub circuit for H-bridge the output from the "MOSFET" equations result in 0 which does not reflect reality.

I need a model that reflects the actual current paths for coasting and brake conditions

Meaning current from a motor coil must go

1) back to VM terminal (via the reverse diodes) for coasting

and

2) circulate in the two low side switches in brake mode.

I desperately need your experts to update the model

NB. The model also lack the deadtime simulation as the datasheet does not specify any information on this,

Thanks

  • Carsten,

    I am sorry that the model is not behaving as expected.  Can you show me simulation results and how you are exercising the inputs?  The outputs should match the inputs per the truth table in the datasheet and support the modes you desire above.  

    Regarding dead time, it is nominally 100ns.

  • Ryan,

    Thank you for the fast reply.

    The truth table states outputs should be coasting if IN1 and IN2 is zero.

    Coasting should turn off both high and low side switches, resulting in a circuit equivalent consisting of the free wheeling diodes and high impedance MOSFETs.

    Simulation below shows that the outputs stay low during coasting, although connected to VM through pullup resistor RA and RB 

    The sub circuit for the HBRIDGE show why the model work like this:

    .SUBCKT HBRIDGE AVDD PVDD VHA VLA VHB VLB VOA VOB SDB SSS VSS
    ESD VSDBX 0 VALUE = {IF(V(SDB)>1.3,1,0)}
    ESS VSSSX 0 VALUE = {IF(V(SSS)>1.3,1,0)}
    EBETA VBETA 0 VALUE = {4.4444}
    GAH PVDD VOA VALUE = { V(PVDD,VOA)*IF(V(VHA,0) >= 1.3,1,0)*V(VBETA,0)*V(VSDBX,0)*V(VSSSX,0) }
    RAH PVDD VOA 100
    DAH VOA PVDD IDEAL_DIODE
    GAL VOA VSS VALUE = { V(VOA,VSS)*(IF(V(VLA,0) >= 1.3,1,0)*V(VSSSX,0) + V(AVDD,0)*(1 - V(VSSSX,0)))*V(VBETA,0)*V(VSDBX,0) }
    RAL VOA VSS 100
    DAL VSS VOA IDEAL_DIODE
    CAH PVDD VOA 25P
    CAL VOA VSS 25P
    GBH PVDD VOB VALUE = { V(PVDD,VOB)*IF(V(VHB,0) >= 1.3,1,0)*V(VBETA,0)*V(VSDBX,0)*V(VSSSX,0) }
    RBH PVDD VOB 100
    DBH VOB PVDD IDEAL_DIODE
    GBL VOB VSS VALUE = { V(VOB,VSS)*(IF(V(VLB,0) >= 1.3,1,0)*V(VSSSX,0) + V(AVDD,0)*(1 - V(VSSSX,0)))*V(VBETA,0)*V(VSDBX,0) }
    RBL VOB VSS 100
    DBL VSS VOB IDEAL_DIODE
    CBH PVDD VOB 25P
    CBL VOB VSS 25P
    .MODEL IDEAL_DIODE D (N = 0.0001)
    .ENDS HBRIDGE

    --------------------------------

    Actually this sub circuit needs redesign.

    This is the most important feature missing in the current PSpice model Rev A,  but it should also be possible to simulate decay mode as this is an important mode also..

    Another feature is the deadtime and (threshold levels on the input, but that is not so important)

    NB.: For the datasheet the deadtime should be specified

    Thank you for your support

  • Carsten,

    I apologize for the error in the model.  You are exactly correct, the current should decay thru the body diodes of the internal FETs.  It appears that the model is incorrectly "braking" instead of "coasting".  

    I also agree that the deadtime should be in the datasheet.  

    Unfortunately, we do not currently have anyone in my team that can fix this model.  

    Would you be able to proceed with your project with an EVM?  These are readily available on ti.com.