ADUM3220 Gate driver floating supply


I have a question about a testing set up that I am implementing using ADUM3220 evaluation board. I have a full bridge of MOSFETs to implement LLC Converter, the input DC bus can go up to 400 v the output after the transformer is 48 v dc and the power is 350 Watt.

I am using three different power supplies (E3630A ) form Keysight, two of them to power the two high side gate voltages and one for the low side switches.

When I start increasing the dc bus higher than 50 v , the gate driver power supply reading (E3630A ) goes up.

I connect the oscilloscope across the power supply but I still can see the constant 15 volt of the gate driver even though the digital reading on the power supply is different.

Any idea on what is happening?


  • Sorry, the gate driver evaluation board that I am using is ADUM3223.

  • 0
    •  Analog Employees 
    on Nov 2, 2015 10:33 PM

    Is this how your circuit is setup?

    I would expect this to work. The bus voltage ground should only connect to the lowside GNDBs. The E3630A should be able to have up to 240 VDC common mode, so 50 V should not be a problem for it.

    One experiment you could do is completely remove the ADuM3223 eval board from the situation, and just bring the E3630A up to 50 V above earth ground, and see if the readout starts changing.


  • Hi,


    Thanks for your reply.


    The schematic you draw is exactly my connection.

    I am using 62050H programmable power supply from Chroma as the DC bus power supply. I think the ground of this power supply is isolated from the earth ground, don't you think in that case I can go higher than 240 V without any damage for the E3630A as the 240 V is specified if the ground will be floating above the earth ground ?

    Just to clarify a bit more, you mean by this experiment that I should connect the positive of the dc bus voltage directly to the ground of E3630A and observe the readout?

    lastly, the wired thing is that during the experiment , if the voltage of E3630A goes up and I connect a voltmeter to measure the output the readout drops again !


  • 0
    •  Analog Employees 
    on Nov 3, 2015 5:24 AM

    "Just to clarify a bit more, you mean by this experiment that I should connect the positive of the dc bus voltage directly to the ground of E3630A and observe the readout?"

    Yes, this would be a useful test. See if bringing the E3630A above earth ground is causing the issue.

    The 240 volt limit is relative to earth ground, but you don't actually know where the Chroma is going to be centered. To guarantee its floating point, you'd need another voltage source between the Chroma's ground and earth ground. One way to simplify your testing is to use batteries to power the high side channels, instead of the E3630A supplies. This will allow you a much higher common mode range, without worry of damaging what is supplying the high side drivers. That way, you can reference the bus voltage to earth ground, which in my opinion is easier and better.

    You can run each ADuM3223 channel off of two 9 V batteries in series, just make sure your gate ringing isn't large (place a 3-5 Ω series gate resistor between the gate driver and the switch being driven). Alternately, you could put an LDO on the two 9 V batteries, and test at some voltage between 0 and 18 V easily.


  • 0
    •  Analog Employees 
    on Aug 2, 2018 3:15 PM
    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