We are using EPC 2304 for our design of 600W 100V to 24V regulated output soft switched buck converter at 1 MHz.
I have checked dev board EPC90140 and it shows operation at 100kHz only so not sure the same gate driver can be used here.
Also, the gate resistors on the schematic/BOM of EPC90140 seem to be too small. The peak output source/sink current of the NCP51820A gate driver is higher than its rated value.
NCP51820 rated output current: 1A source/2A sink.
EPC2304 Vgs: 5V
Internal gate resistance of EPC2304: 0.4ohm
Gate resistors on the EPC90140 R80-R82 : 0.47 ohm.
Thus, current seen by the gate driver: 5V/(0.47 + 0.4) = 5.7A which is much higher than rated value.
Hello,
for the gate driver selection I would start by focusing on two criteriae:
1- can the gate driver deliver enough current (is it low impedance enough) to charge the Ciss (combination of Ggs and Qgd depending on exactly your topology and conditions) to achieve the required transition times?
2- can the gate driver dissipate the energy associated with charging and discharging the Ciss + its internal circuitry consumption and stay within an acceptable temperature
My recommendation is to first implement your circuit in Spice (you can find models for our devices here: EPC Device Models (epc-co.com) (LTSpice being our main tool)
For the power dissipation the gate driver also has several current consumption charts as a function of switching frequency that should be able to guide you in the design.
In terms of the gate resistors on EPC90140 they were chosen based on the gate switching form to optimize the trade-off between switching speed and ringing. Note that the current outputs in the gate driver datasheet are given as a typical, and are a way to specify the impedance of the gate drive circuit. The limitation will be the thermal dissipation in the device, which is related to the gate charge and not the switching current.