Can the UCC28180 directly drive an IGBT ??

The FGH40T100SMD IGBT appears to need atleast VGE=15V in order to deliver the lowest VCE,ON in the on-state. In fact applying a higher VGE=20V appears to reduce VCE,ON voltage drop even further, running the device in deeper saturation and reducing the on-state power losses.

UCC28180 gate output voltage is clamped at typical of 15.2V, but due to part to part variation the clamp voltage can be as low as 14.5V (min) or as high as 16.1V (max). If the UCc28180 is used to drive the IGBt directly, then please make sure that the VDD bias applied to the UCc28180 device is high enough that will ensure that the full gate drive is available to drive the IGBT.  VDD=18V would be a good choice, even higher would be better but need to make sure that the bias voltage is ALWAYS no higher than the recommended maximum of 21V for the device.

Also, please note that the VCC,ON threshold of the UCC28180 device is 11.5V Typical, which means that when the VDD bias applied to UC28180 exceeds 11.5V, the part will start delivering gate output pulses of similar magnitude. At this time the IGBT gate drive voltage will also be around 11V and there will be a lot of conduction loss in the IGBT in the first few pulses. The key message here is that the VDD bias supply to the device must have a fast ramp rate between the VCC,ON level and the final intended level of 18V.

Here is the simplest solution: We would highly recommended to use a dedicated gate driver device such as UCC27531 to drive the IGBT. This device can accept VDD voltage as high as 20V and can comfortably drive the IGBT into full saturation in the onstate and deliver the lowest conduction loss. Whats more, if needed, the IGBT can be switched off with a negative voltage as shown in Figure 10 of the UCC27531 datasheet, using the special biasing arrangement. This approach is commonly used to ensure more robust design and higher noise immunity, especially in high power applications. The negative gate voltage in the off-state also delivers immunity against Miller-induced C*dV/dt turn-on.

Finally, please note that IGBTs have a long tail current at turn-off. This means enough OFF time must be allowed to allow the IGBT tail current to go to zero in the off state. The tail current is longer and hence a longer off time is needed when the IGBT is running a higher magnitude of the current or driven deeper in saturation. In PFC application, the on-time duty cycle near AC zero crossing is nearly 100%. For example, UCC28180 is able to deliver 96.5% typical. This means only 3.5% of the switching cycle is available at turn-off. The key point here is that the switchign frequency must be low enough to allow the tail current to die down to zero. Please consult with the IGBT manufacturer to establish how much off time needs to be provided to ensure the tail current goes to zero. The programmable frequency feature in UCC28180 will allow you full flexibility to choose the appropriate swtiching frequency needed for the kind of IGBT you have chosen.

Keeping in mind all the points mentioned above, please ensure that the IGBT is not running too hot in your application. This may very well be the application issue you are encountering. Good Luck !