2SC0435T status output not reporting fault when the driver is damaged

Posted by: GasperHribar on 06/09/2024

Hello,

We are experiencing an issue with the status outputs (SO) of the 2SC0435T when the driver is damaged. Specifically, faults are not being reported, and the SO is always in a high impedance state, even though the fault criteria (secondary side under voltage or overcurrent) are met. By "damaged," I mean that the DC/DC MOSFET is burned out or destroyed, and the LDI on channel 2 is also destroyed (see the attached picture).

Is it normal for the driver to not report fault status in such a state? How do you recommend detecting such damage?

Files

Attachment	Size
20240609_155635.jpg	3.49 MB

Comments

Hello

thanks for approaching us.

Here is the feedback

The UVLO function is expected while the driver core is in healthy status, which can report under voltage warning for the primary side and secondary side supply voltage.

However, the driver is severely damaged in your case, it can not report any failure signals.

According to the attached picture, the driver experienced severe electrical over stress condition, causing damage. one possibility is overvoltage on the driver, which can damage the driver without having under voltage, in this case , the driver will be not be able to report feedback signal, since it had been damaged already before the voltage dropping is coming

Thank you for your answer.
So there is no way to detect that the driver is damaged without doing visual inspection. This is extremely problematic for us as we are driving two parallel IGBTs on one full bridge leg. Without knowing that one driver is damaged all of the current is going over one IGBT instead of two and then the destruction of other IGBT is almost certain.
The attached picture represents the driver after IGBT is blown up/destroyed, for example collector get shorted to the gate.
Do I understand correctly that in such scenario fault will not be reported?

How do you do the IGBT parallel ? In normal case, if a core is damaged as your showed, all paralleled IGBT should stop work.

Another driver is driving the parallel IGBT module, we of course have added inductance between the two IGBTs, to address the possible difference of turn on times.
So the damaged driver is driving upper and lower IGBT in the full bridge application and there is a second driver which is driving the other two parallel IGBTs.

Hello, I have further question about SO. Is it expected to not get fault if the IGBT is not connected. We were previously using IHD660 and we got fault if the IGBT module was disconnected during ON scenario (generated pulses on input pin INA/INB).

Hello, I still haven't received an answer for the latest post. To better describe the upper scenario I am talking about disconnecting IGBT when there is no DC link voltage applied to the IGBT.