Other Parts Discussed in Thread: LM34936
Our initial development we used the development board (BoostXL-DRV8323RS) to drive our motor successfully and have not seen any issues.
Our production board uses the DRV8323S and is powered via an LM34936 buck/boost chip. The DRV8323S showed a fault when we toggled the Enable line. With no PWM on the input to the DRV8323S, we power up with the enable held low. When we enable the DRV8323S there is a very brief clunk from the motor and the power to the processor dips for long enough that it resets. We then go through the startup process again, and the fault repeats.
All we are doing is providing power to the DRV8323S, we are not trying to drive the motor but it seems that raising the enable line generates a very brief transient of 175A in the motor (measured by monitoring an overall current sense resistor with an oscilloscope), this trips the protection in the LM34936 and it goes into hiccup mode, hence the loss of power to the processor.
We have been able to proceed by disabling the hiccup mode on the LM34936 but as this provides an extra degree of protection over and above the cycle by cycle protection it is not our preferred route and we are not able to provide the 16ms of holdup that would be required to prevent the processor powering down and allow us to reinstate hiccup mode.
This prompts some questions:
Do you have any ideas as to why simply enabling the DRV8323S should cause this transient?
Are there any things we should be doing to the configuration to ensure the chip starts up in a ‘safe’ way to avoid pushing energy through the output FETs? E.g. should we be applying PWM before enabling the chip?
All suggestions gratefully received
David Lambert
