IPM(Intelligent Power module) connection

[seyyah]

intelligent power module ipm 15

This is my first use of an IPM. My purpose to use this igbt module is to simplfy the circuit but i'm a bit confused. I think it doesn't help to simplfy the circuit. Look at the attached schematic and help me about the connection please. The datasheet is here:
http://www.fujielectric.de/SEMI/Datenblatt/IGBT/6MBP50RA-120.pdf

My questions:

1- Can we directly interface this module to the mcu? If so all grounds will be connected to the logic ground.
2- If we must use optoisolator then can we use one supply for high side?

 

[Guest]

You must use optocouplers because connecting grounds together at upper part of the bridge means making short circuit between the output phases.The same thing is walid for using the same ( or one ) voltage for all the upper parts of the bridge.If you want to use just one voltage for the upper part of the bridge tray bootstrap circuit ( it has the limitations of the driving current and the switching frequency).
Greetings
Dragan

 

[seyyah]

Ok, i formed 4 isolated 15V supply. Connected all pins except gate pins. Didn't apply dc bus voltage. I tried to make to conduct the igbts by applying 5V to their gates which is obtained their own supply by resistor divide. But none of the igbts went into the conduction. So what may be the problem? I don't believe that driving an igbt is a very hard thing. But i always failed to make it work properly. The reason that i choose an ipm is to eliminate my inexperience but i couldn't make it work also. Is there a curse on me?

 

[Guest]

Usualy you need 15VDC (9-15) to drive an IGBT (to bypass the miller zone of the input FET ). If you do not bypass the Miller zone you can A) not open the IGBT if the voltage is to low (what happened to you) and B) work in the linear mode ( which can burn the IGBT and C) you can open it succesfuly.To work in mode C try to 1.) not supply the high voltage 2.) Switch the driver on and monitor the drive signal at one chanel (you must see the Miller zone - change in slope and be sure that you have surpassed that point for at least for 3-4V). Try to check if you have ringing in driving signal and measure the attacking slope of the signal ( must open in 4-10uS ) if the time is longer means that you do not have enough driving current for the IGBT or the gate resistance is to high ).

Hope this had helped a little bit.

Greetings
Dragan

 

[seyyah]

Did you check the pdf? IGBTs are on when input is 0. And off when input is 1. The input zener is at 8V. From 1.95V and above the igbts are off. Below 1.5V, they are on. This module has driving circuit, that's why the inputs are as low as 8V or lower. The driver inside, applies 15V to the gates. This is an IPM module, not an igbt module.

 

[Guest]

To be onest I must say that I did not checked PDF and that I did not use IPM-s. I usualy work with HCPL316 or TLP251/TLP521 depending on the driving request and of course the price of the equipment. I have tested IR2112 and 2136 in some not to demanding situations.


Greetings
Dragan

 

[dualcorecpu]

Ok, i formed 4 isolated 15V supply. Connected all pins except gate pins. Didn't apply dc bus voltage. I tried to make to conduct the igbts by applying 5V to their gates which is obtained their own supply by resistor divide. But none of the igbts went into the conduction. So what may be the problem? I don't believe that driving an igbt is a very hard thing. But i always failed to make it work properly. The reason that i choose an ipm is to eliminate my inexperience but i couldn't make it work also. Is there a curse on me?

You mentioned that to turn on the IGBT, the Vin = 1.5V or lower. But you apply 5V???
Do you think we can turn on a IGBT without forward bias C-E or without DC bus voltage?

 

[seyyah]

The problem is i think i pre apply pulses and continous pulses. The gate driver must see falling edges to make igbts on.