Gate Driver Circuit
The layer between the logic output and the gate of a power MOSFET: it charges the gate capacitance with an amp-level pulse and shortens the transition.
A gate driver circuit is the intermediate layer that stands between the milliamp-level logic output of the microcontroller and the gate of a power MOSFET. Its job is to charge and discharge the gate capacitance with a pulse current in the amp range, and so to shorten the transition time of the switch between conduction and cut-off. Typical integrated drivers give 1–4 A of peak current and bring the transition down into the 50–300 ns range.
Electrically the gate is a capacitor
The gate of a power MOSFET looks like a capacitor of a few nanofarads. The faster you charge this capacitor, the faster the MOSFET turns on. A gate connected directly to a microcontroller pin stays in the partial-conduction region for microseconds because of the pin's 20–40 mA limit, and during that time the switch heats up like a resistor. The amount of charge that has to be delivered is given in the data sheet as the gate charge (Qg), and the peak current the driver has to supply follows from that value. As the gate charge grows, that is as the MOSFET grows, a stronger driver is needed for the same transition time. Using a large switch in a high-current module and leaving the driver small is the typical reason behind designs that look right on paper but heat up in the field.
The gate driver does two jobs at once
- High pulse current: It charges and discharges the gate quickly and shortens the time spent in the partial-conduction region.
- Above-supply voltage: For the high-side switches to stay in conduction, a gate voltage above the source voltage is needed; this is usually produced by the bootstrap method.
A fast transition is not always good
Shortening the transition lowers the switching loss but raises the rate of change of current and voltage. The price of that is electrical noise, the voltage overshoots that appear when it combines with cable inductance, and the opposite switch in the bridge leg turning on unintentionally. The designer strikes this balance with the gate resistor: as the resistance grows the transition slows down, the noise falls and the loss rises. The right point varies with the switching frequency of the application and with the cable arrangement; the setting that looks most efficient in the laboratory can be the noisiest one in the field with a long motor cable. For the same reason a dead time is put in between so that the bridge legs do not conduct at the same moment.
It is not an area for the user to intervene in
This layer is inside the module; it is not seen by the user, not adjustable and not replaceable in the field. When a heating problem occurs on a driver module, the answer is not to intervene in the gate driver but to lower the current, to ventilate the installation or to move up a model. Changing the gate resistor or upgrading the driver chip upsets the protection and EMC balance of the board: when a single component changes, the transition speed, the noise and the dead time all shift together. The warranty is void in the case of such interventions as well.
What it means in practice
This layer is invisible but its results are visible: a power stage that is weakly driven heats up far more at the same current and cannot withstand a high switching frequency. A driver module's ability really to carry the continuous current it declares depends largely on this layer being designed correctly. That is why two modules that state the same current rating do not behave the same way in the field; the difference lies in this design layer, which does not show in the catalogue. When choosing a module, read the declared current not on its own but together with how much it heats up at that current and how long it can carry its peak value.
The context in which this term is used: DC Motor Driver Selection Guide