ISO5852S: desaturation inactive higher than 1V

Hello Harrie,

At first glance it seems that there may be some reverse current due to the IGBT or MOSFET turning off that is causing the capacitor to charge to above the threshold voltage. One possibility would be to increase R226 to 1 kOhm. What is your target voltage threshold for DESAT detection and your desired blanking time? Is is also necessary to use R222?

Does the DESAT show normal voltage waveform during turn on? Could you share that waveform? 

Regards,

Audrey

Hi Harrie,

In any case, the output is not affected since the driver is turned off. You also should not be seeing the nFLT pin because the driver input is low. Is that correct?

Regards,

Audrey 

Hello Audrey,

Our proto inverter is with SiC mosfet. 
{700 V DC, 40 kHz, sine-filter}

The target threshold is now 3.5 V 10 us (R222 is necessary @ 220pF)

We started this measurement because after 30 minutes at 80 kW we get a Fault signal from the TI-driver.
the SiC-mosfet temperature of that phase was ca. 50 C. Our electronics above the SiC-mosfet was in free air (21 C).

We did the measurement at 0 kW (sine-filter connected) and a "cold" system and we did not have a Fault signal at that moment.

Best regards,

Harrie

Hello Audrey,

waveform turn off / turn on.  I don't see any problem in turn on, only the desat-unload-capacitor mosfet can do some more current at the start of the pulse?

I have this picture also in excel measurement (500MSa/s)

Regards Harrie

Hi Harrie,

Thanks. I see how you have come out with 3.5V DESAT threshold voltage based on your schematic. However, a 10 us blanking time seems much longer than what you would actually get, which is more like sub 2 us blanking time with 220pF cap. But since you are using SiC, then I would expect a sub 2 us blanking time is acceptable given the fast turn-off speed.

Thanks for sharing the additional waveforms. It seems to be an issue then with the high positive dv/dt of the SiC MOSFET during turn-off. If possible, could you try increasing R226 to 1 kOhm to see if there is any improvement?

Do you have a way to measure the current at the DESAT pin during this transient?

Regards,

Audrey

Hello Audrey,
after some struggling I have measurements (not from current).
I did the measurement  with 100E and with 820E, extra measurement channel 1 U drain-source.
picture 1: 100E trigger on high Desat signal.

picture 2 100E lower Desat signal

Picture 3 820E try to trigger on highest desat signal (when turned off)

picture 4 820E On-Off puls

The high Desat pulse is gone.

To understand the measurements is it the high impedance of the desat fet in the ISO5852 with the high dU/dt peak the problem of the signal? and is the ISO5852 blocked for desat fault trips when Gate signal is low?

There will be one challenge left for us and that is to speed up the desat trip time.

We think the R222 must be changed from 10k to 1 k, can You give me some feedback for the two questions and this last statement.

regards Harrie.