IGBT mimum on-time requirement 1

[DHambley]

Does anyone have any explanation showing why there is a minimum on-time of an IGBT? My colleague said we need to have a minimum because he heard it long ago. All I found on a google search is an app note from Semikron saying that the effect of turning off an IGBT when it is still turning on is that there will be an increased di/dt (because the PNP portion isn't saturated yet) and so this larger di/dt could result in a higher inductive voltage spike at turn off. The app note does not give any curves of this effect vs gate resistor, temperature or current or anything.

Have seen any other articles explaining how to plan for this minimum on-time effect? Thanks.

 

[crthompson]

This literature gives a formula to calculate your minimum on time...
Hope it helps...

 

[iop995]

There are two reasons that request a minim on-time; first - like any real element, need some time to do something (there are not zero response time real device), so to turn-on IGBT need an amount of time. Second, minim on-time may be requested by converter topology.

 

[electricuwe]

There are a lot of reasons, why the PWM modulator should ensure a minimum-on time. Some are related to device physics, some are related to the driver circuit, some are related to protection and some are related to the load.

In regard to device physics, these two issues are most relevant:

- minmium on-time of the IGBT is also minimum-off time for the diode
- if an IGBT is turnend off immediately after turn-on carrier modulation has not fully taken place and the device behaves like a Mosfet, this especially means that the device turns off faster, causing higher overvoltage. Do some double-pulse testing with decreasing on-time and you will see that for otherwise identical conditions overvoltage will be higher, if the IGBT was only conduction for a very short duration. This effect of course is different for IGBTs of different blocking voltages.

The reference provided by crthompson gives the minimum-on time requirement arising from a specific driver function - two-level-turn-off - and is not directly related to the characteristic of the IGBT itself.

 

[electricuwe]

Small correction to my previous post: carrier modulation should read conductivity modulation