AduM4135 Gate Ringing

Fig1 AduM4135 Gate Ringing

Fig 1 AduM4135 Ringing

Fig2 ISO5451 

Hi, I'm using the Adum4135 at 20kHz with Rgon=Rgoff=4.7Ohm. I noted a ringing problem in the gate turn on(see fig1) The turn on is ringing at about 30MHz. At the beginning I thought that the problem was in my pcb layout but when I replace the AduM4135 with the ISO5451 from TI, the ringing problem disappear (see Fig. 2) I saw the datasheet and I think that the AduM4135 and ISO5451 have the same switching behavior and they are also pin compatible (at less of 180° of rotation). So, why there is this ringing problem in the AduM4135?


  • 0
    •  Analog Employees 
    on Nov 25, 2018 8:47 PM


    Thanks for your question. From the waveforms, it looks like the total voltage swing on the test performed with the ADuM4135 is just over 15 V, while the test performed with the ISO5451 swings to across just over 20 V. Are these two waveforms performed on the same board, and same secondary power supply? It looks like there is a difference there in test conditions.

    In the ADuM4135 test, it almost looks like either the plateau voltage of the switch has shifted, or the probe is in AC mode.

    Are you able to share any parts of the schematic, or pictures of the layout?



  • Hi,
    thanks for the answer. In Fig1 there is the a part of the schematic. Vdd2=15V (between pin 13 and pin 11) and Vss2=-8V (between pin 16 and pin 11). The fig doesn't show the secondary voltage schematics. It is a solution with resonant low cost ZVS. However I think that I found the solution of my problem. I think infact that the problem was in my test equipment. I'm using a P5200A Tektronix differential probe with BW=50MHz and Cin=2pF. Also I used two Tektronix extender leads of about 1.5meter each. This extender have about 10uH of inductance. The resonance in my measure path is about 35MHz. The rise time of the Adum4135 is 12ns and the rise time of the ISO5451 20 ns. If I use a rule of thumb of the bandwith of a signal from its rise time (BW=0.35/tr) I have for the Adum BW= 29MHz (near the mesaure path resonance) and for the ISO5451 BW=17MHz (far from the measure path resonance). If I don't use the extender the ring problem disappear. I think this is the explanation.

  • 0
    •  Analog Employees 
    on Nov 26, 2018 5:39 PM in reply to Prospy

    I'm glad you found that. Measuring gate drive waveforms is definitely an interesting topic, and I'm learning more all the time. The easiest solution I've found for measuring the highside gate waveform is to use Tektronix ISOVU probes. They are really costly, but work really well for differential high bandwidth measurements.