LM74670-Q1: Question of LM74670

Hi Jimmy,

MTE050N10KRV8 is a 100V rated FET and hence there should not be any failure due to over voltage applied across its drain-source for AC input: 16-24Vac. The Vgs rating of the FET is also high enough.

Reverse current flow should not be a problem as long as the FET is enhanced where the voltage across the FET is very low.

The only probability I could think of is FET SOA violation. Can measure the gate-source voltage applied to the FET and share the waveform captures.

Hi GD,

We appreciated your great support. Please refer below,

This waveform has been measured to follow Vgs off and then the current returns to 0, please confirm whether it is a normal phenomenon (picture 2)?

Many Thanks,

Jimmy

Hi Jimmy,

Let me review and get back to you by tomorrow.

Hi GD,

Thanks for your great support.

I added some new waveform for your reference.

This is the waveform of the measured mos Vgs and current. It can be seen that the current will reverse for a period of time before turning off the MOS as below,

Many Thanks,

Jimmy

Hi Jimmy, Thanks for sharing additional waveforms. Let me check and get back to you.

Hi GD, 

Thanks! Please let me know if you have any recommend on this.

Hi Jimmy,

The current through the FET is expected to be in reverse direction for a short duration before the FET is turned OFF. This is not an issue as the FET is fully ON and it will allow bi-directional current without any issue .

In the waveform below I can see that the Vgs signal drops to negative occasionally, this is not expected. Is there any offset added to the Ch4 measurement ?

The FET Vgs is expected to be as shown below,

Hi GD,

Thanks for your quick support.

According to the datasheet, the Vgs is negative when charging the charge pump cap.

The waveform I showed is the time as the following point.

(from tidt194)

Many Thanks,

Jimmy

Hi Jimmy,

The waveforms look okay to me and I do not see any reason why the FET should get damaged.

It will be helpful to find the root cause may be if you know the exact test condition during which the FET damages. Is it a repeatable condition ? Do you have waveform captures at the instant of FET failure ?