Snubber for ADuM447x

The UG-492 document on the ADuM447x evaluation board on page 4 states "R4, R3, C4, and C5 are provided for an optional and unpopulated snubber, which can be used to reduce radiated emissions." The part seems perfect for my isolated DAQ application and works great in my intended circuit, but the eval board has about 600 mVpp of nasty 20MHz ringing on the ISO GND (P8/P10) with respect to input GND (P2/P4/P6) and will probably kill my CE radiated compliance. Can you provide further guidance on a suggested snubber circuit or design method? I don't know how to estimate either the effective capacitance or inductance for classic snubber design methods.

I am using 5V in, +/-5.5V out at 900 kHz with the Halo Electronics TGRAD-560V8LF transformer which fits my physical space.

Thanks for all suggestions.

Parents
  • Sorry,

    Project is a revision of a previous isolated DAQ card intended for measuring line voltages up to 277 Vac and shunts down to 50 mVac. The ADuM447x is very attractive with its excellent creep distance and withstand voltage.

    Positive load is 150 mA max, negative is 50 mA. Load is pretty constant except positive can decrease to 120 or 90 mA if two 30 mA relay loads are both off. ADuM output voltage is set to +5.5V and the unregulated negative tracks quite well at -5.7V. Both rails are followed by LDOs and the DAQ circuit (AD8220 inamp, 200kHz AAF, 1MS/s 16b ADC, isolators) runs from +/-5.0V. Schematic of the load circuit is available privately if needed. (An ADuM4474 4:0 isolator will be used for a low-speed SPI port that controls the front end settings, but that function is currently filled by an Si8640 on my prototype card.)

    One more bit of background info. If my load is removed completely and I measure the EVAL-ADuM4070EBZ board alone from IsoGND to input GND, the waveform I sent yesterday is similar in nature but greater amplitude of ~1000mV and lower duty cycle. No surprises there: the lump capacitance of my 8 ch of isolators back to earth ground reduces the P-P voltage.

    Thanks,

    Gary

Reply
  • Sorry,

    Project is a revision of a previous isolated DAQ card intended for measuring line voltages up to 277 Vac and shunts down to 50 mVac. The ADuM447x is very attractive with its excellent creep distance and withstand voltage.

    Positive load is 150 mA max, negative is 50 mA. Load is pretty constant except positive can decrease to 120 or 90 mA if two 30 mA relay loads are both off. ADuM output voltage is set to +5.5V and the unregulated negative tracks quite well at -5.7V. Both rails are followed by LDOs and the DAQ circuit (AD8220 inamp, 200kHz AAF, 1MS/s 16b ADC, isolators) runs from +/-5.0V. Schematic of the load circuit is available privately if needed. (An ADuM4474 4:0 isolator will be used for a low-speed SPI port that controls the front end settings, but that function is currently filled by an Si8640 on my prototype card.)

    One more bit of background info. If my load is removed completely and I measure the EVAL-ADuM4070EBZ board alone from IsoGND to input GND, the waveform I sent yesterday is similar in nature but greater amplitude of ~1000mV and lower duty cycle. No surprises there: the lump capacitance of my 8 ch of isolators back to earth ground reduces the P-P voltage.

    Thanks,

    Gary

Children
No Data