• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 1121 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AMC1301: Switching noise conduction

Gokhan Ozcetin
Prodigy 105 points

Part Number: AMC1301
Other Parts Discussed in Thread: AMC1200

I am using AMC1301 for in line phase current measurement. Since its input stage differs from AMC1200, the datasheet does not offer an input filter but an output filter.

However, the switching noise is more or less conducted even through the first order low pass filter at the output (1k series with 47n to GND). I added independent filters to both outputs of AMC1301.

Is it possible to add a filter to the input stage to further reduce the noise? May be a low value resistor (on the order of 100 ohms) with a large cap?

  • 0
    TI__Guru 61272 points
    Hi Gokhan,

    Thanks for your post!

    Adding a filter to the input stage is fine and 100ohm for the resistor will work. We recommend using a differential cap between pins 2&3 as opposed to single ended implementation.
  • 0 in reply to Alexander Smith
    Prodigy 105 points

    Hi Alexander

    Thanks for the quick reply. Considering shunt measurement applications, is there a limit for the filter resistances  (due to differential amplifier input configuration maybe)? Will it change gain? What is the accuracy of the input stage

  • 0 in reply to Gokhan Ozcetin
    TI__Guru 61272 points

    The differential input impedance is only 22kohm so I would try to keep the resistance fairly low to avoid a large mismatch in resistance due to tolerance variation and increase the size of the capacitor to adjust the cutoff frequency. Signals that have a higher frequency than the cutoff frequency of the filter will be attenuated while signals below the cutoff frequency will be unchanged.

    What is the frequency of the signal being sampled and what is the frequency of the switching noise?

    The cutoff frequency will be 1/(2*pi*2*R*C).

  • 0 in reply to Alexander Smith
    Prodigy 105 points
    Well the pwm is 17.5 khz and the phase current will be a sine wave having frequency of 3khz maximum.

    I added the first order filter having a 3.3khz cutoff to the output. Unfortunately it seems unadequate. I will remove output filter and place an input filter with the same cutoff frequency. Any further advice?
  • 0 in reply to Gokhan Ozcetin
    TI__Guru 61272 points
    I see. With the current 3.3kHz filter you should see about -15dB attenuation for the pwm noise, and around -2dB for your 3khz sine wave.

    Moving the filter to the input should help by not passing the noise through the gain stages of the device, however since the frequencies are so close I'm not sure if -15dB will be enough to meet your criteria. As you pointed out, there will be a further reduction in gain due to the increased resistance on the inputs, but it should be minimal (~1%) if your input filter stays at 100ohm + differential 240nF.
  • 0 in reply to Alexander Smith
    Prodigy 105 points
    So it is best to both apply the filter and increase the pwm frequency then. Thanks for all the information.

    I believe it is better to include such a filter application in the datasheet. The absence of such information made me feel that input filter is not applicable for amc1301 since it was given in the datasheet of amc1200 but not in that of amc1301
  • 0 in reply to Gokhan Ozcetin
    TI__Guru 61272 points
    You're welcome Gokhan! I agree, I'll make a note for the writer.