The power sequencing requirements can vary between different MOSFET driver ICs, so it's crucial to consult the datasheet or user guide of the specific IC you are using. The manufacturer provides guidelines on power sequencing to ensure proper functionality and prevent potential damage.
Control section power and signal come first, else you have no
authority over power stage activity (unless a well-designed
UVLO includes front end status). Ideally "sit there and shut up"
is the outcome of control inputs grounded, so a safe place to
apply control-section power without needing those signals to
co-ramp or input protection to allow over-rail during ramp.
Control section power and signal come first, else you have no
authority over power stage activity (unless a well-designed
UVLO includes front end status). Ideally "sit there and shut up"
is the outcome of control inputs grounded, so a safe place to
apply control-section power without needing those signals to
co-ramp or input protection to allow over-rail during ramp.
So, you are saying the MOSFET gate signals should come first before MOSFET drain voltage is applied, right? Is it common to all MOSFET driver ICs across all manufacturers?
The obvious requirement is to keep MOSFETs in safe off state during power-on. It can be achieved in different ways.
Powering Vdd before Vin is specified for the MIC4606 development board. The Application circuit given in the datasheet however derives Vdd from Vin and respectively doesn't strictly obey this rule. But it's only a principle schematic that doesn't show details like voltage converter startup parameters.
Additionally, you need to consider that the high-side of a bootstrap driver is never powered before the low-side starts switching and doesn't keep powered unless the low-side is periodically activated.
You are reporting that MOSFET driver was damaged due to wrong power sequence.
1. The driver gets rarely damaged on it's own. In most cases, blown MOSFETs with drain-gate short are causing the damage.
2. Datasheet states that driver outputs are kept low as long as the UVLO threshold isn't exceeded. So even with wrong power sequence, the driver shouldn't cause output shoot-through.
3. I would also consider other causes of failure, e.g. uncontrolled switching due to bad circuit layout or uC software.