• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 48 subscribers
  • Views 1774 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.

Anti-aliasing for ADS1298

Mohith Manoj
Intellectual 290 points

Other Parts Discussed in Thread: ADS1298

The schematic for ADS1298ECG-FE has a Radio Frequency Interference filter at 153KHz and no separate anti-aliasing filter. What exactly is the criterion for selecting the low-pass filter at the input of the ADC.

The datasheet of ADS1298 details the digital sync filter where it is mentioned that the RC filter at the input should be capable of attenuating frequencies Fmod and its multiples. While operating with 2.048MHz clock and High-Resolution mode the Fmod stands at 512KHz.

Is it sufficient that we have an RC low pass filter on each channel that can sufficiently attenuate 512KHz? If so, what is the level of attenuation desired?

  • 0
    TI__Guru**** 182761 points

    Hi Horizon,

    It's been a little while since we talked.  How have you implemented the bypassing on your board?  Have you gotten past the lead off detection issues?  This is a great question regarding the ADS1298!  I'm sorry for the delay but we should be able to give you more specific details regarding the question(s) above tomorrow morning. 

  • 0 in reply to Tom Hendrick
    Intellectual 290 points

    Dear Tom,

    Nice to hear from you.

    For the bypass capacitor configuration, I took your advice and installed 100nF X7Rs on all the supply pins with common 10uF tantalum bulk capacitors. With this I am able to read with the input multiplexer set to 'input short' and get a noise level close to 4~6uV. For clock and reference I am using on-board features.

    I got the board running yesterday only and so far I have read normal input data, noise data, test waveforms, supply and temperature only. I am yet to try lead-off detection.

  • 0 in reply to Tom Hendrick
    Intellectual 290 points

    Tom,

    Lead-off detection works like a charm... I am using current source mode.

  • 0
    TI__Expert 3260 points

    Horizon,

    The first main question here is whether or not you're using PACE, or if you're just concerned with ECG. Upon design of the EVM, most customers requested a -40dB/dec at the 512KHz mark (hence the 2 RC filters), because it is at that frequency that you're worried about the sinc filter's foldback (you may look at page 29 for further reference). If you are only concerned with measuring ECG and not concerned with PACE, you can design the antialiasing filter with an earlier roll off, perhaps 10k or 100k, depending on the signal components you are interested in.

  • 0 in reply to Michael W
    Intellectual 290 points

    Michael,

    Thank you very much for the clarification. We are not planning for PACE detection now. We have a 22KHz RC for RFI protection. I guess that would suffice for the present deign.

  • 0 in reply to Mohith Manoj
    TI__Mastermind 43915 points

    Hi Horizon,

    That should be fine. Some customer's are interested in some of the details of the PACE signal which are at higher frequencies which is why our anti-aliasing is so far out. We wanted to make sure to pass as much of the signal through as possible. If not interested in those higher frequency components 22KHz should be fine. Especially considering a large part of the PACE signal components are within the 1KHz - 10KHz frequency spectrum.

    Regards,

    Tony Calabria

  • 0 in reply to Tony Calabria
    Prodigy 10 points

    Hi all,

    If one is only interrested in basic ECG infromation (no pace), can the AA filter be set even much lower, say about 150Hz ?

    Regards,

    Pierre