Post Go back to editing

ADUM3223 blanking high side for 200uS periodically

I am using the ADUM3223 with EPC2012C devices switching at 1MHz.

When rising input voltage of the converter leg above around 13V, the high side driver blanks the output for 200us periodically ( period of kHz). Output pulses are down from 5.0 to around 0. something volts , causing the mosfet to not turn on and the inverter to fail.

I populated another board, and after some OK running time the ADUM failed in similar manner. 

But in open loop tests, i have kept the system running OK for days. 

Any idea about any protection / issue which might occurr ? or failure mode of the device? 


  • Hello,

    200 µs is very long, and we haven't encountered anything of that length before. I don't know what could be causing a delay that large. The largest possible delay that I'm aware of is the power up delay, which is around 50 µs maximum (found in the logic table of the datasheet).

    If the part is working open loop, but failing closed loop, is there a chance the controller is the difference? Have you monitored the VIA and VIB input pins?

    Are you operating the system in a bootstrap configuration? Could the highside be losing power at low duty cycles?

    If you are able, could you share a partial schematic? If you need to keep the schematic out of public forums, we can email if needed.


  • Yes, it was a bootstrap issue. The capacitance value had to be adjusted, and the issue has been resolved. 

    Another question i have is: What is the maximum switching frequency for reliable operation of this  component?

    The 1MHz is the absolute maximum of the chip or it is calculated from thermal dissipation depending on the gate capacitance being driven?

  • Hello,

    The 1 MHz comes from expected power dissipation in normal operation as you guessed. The ADuM3223 is capable of faster operation. The true limitations come from power dissipation and minimum pulse width. In our production testing, the quiescent current limits are set to a maximum frequency of 1 MHz. If the switching frequency is increased above 1 MHz, the datasheet limits for the quiescent currents will no longer hold.

    The minimum pulse width of 50 ns is the shortest you should run pulses (both in the positive or negative directions in order to get a proper waveform out of the gate driver. As long as power dissipation is OK, you should be able to run at very high frequencies well above 1 MHz. This is operating outside of datasheet bounds, though, so exact performance is not guaranteed.


  • This question has been assumed as answered either offline via email or with a multi-part answer. This question has now been closed out. If you have an inquiry related to this topic please post a new question in the applicable product forum.

    Thank you,
    EZ Admin