• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 1340 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

ADS1294 antialiasing filter when not using the RLD electrode

Cameron Elvey
Prodigy 115 points
Other Parts Discussed in Thread: ADS1294, AFE4300

Hi, so I'm trying to figure out a system where I'm using the ADS1294 with only two electrodes and no reference electrode and I'm a bit confused about the filtering needed (which is stressed as being important to get right). I'll be using it in full power mode where the Nyquist is said to be 256 kHz but then there is the digital decimation filter as well.


I'd prefer to do filtering in the digital domain with all things being equal too so just want to avoid aliasing without losing the signal. Is it possible to get any suggestions as I'm at a bit of a dead end? Thanks

  • 0
    TI__Mastermind 38745 points
    Hey Cameron,

    The input PGA on this device actually has low pass characteristics (see table 5 in the datasheet). Besides that, most folks just use an RC filter for anti-aliasing. Like you mentioned, the delta sigma samples fast, low pass filters, and decimates so your risk of aliasing is reduced and the transition of the anti-aliasing filter doesn't need to be too steep.

    Beyond an anti-aliasing filter, you will also need some way to maintain the input common mode within the measurement range of the ADC. In systems that use an RL electrode, the ADS1294's RL drive amplifier can be used to bias the patient. In your system, however, I would recommend simply ac coupling the inputs. That is to say you should place a dc blocking cap in series with the inputs and use a simple voltage divider with large value resistors to bias the inputs at the device. Be careful about the values you choose, though, since that type of setup is actually a high pass RC filter, so you'll have to take its frequency response into consideration. There is a treatment of this topic in section 10.1.2 of the datasheet.

    Regards,
    Brian Pisani
  • 0 in reply to Brian Pisani
    Prodigy 115 points

    Thanks for your reply Brian,


    The Figure 95 in the section 10.1.2 is partly what is creating the confusion, sorry I should of referenced it in my original question. It sill has the anti-aliasing/protection blocks shown and I'm unsure of their performance requirements. Sorry if its a bit of an elementary question.

    Cheers,

    Cameron

  • 0 in reply to Cameron Elvey
    TI__Mastermind 38745 points

    Hey Cameron,

    I drew that diagram. I the reason I just put a block there rather than put specific components is that 1) I didn't want to suggest how to do patient protection because I am not an expert there and 2) some designers will put a cascaded two-pole RC filter instead of just a single RC. You, of course, may do as you like, but remember that there is a tradeoff between band limiting and common-mode rejection as is explained in section 10.1.3. I like to use TINA-TI to simulate different anti-aliasing schemes.

    Regards,

    Brian Pisani

  • 0 in reply to Brian Pisani
    Prodigy 115 points

    Hi Brian, (second go at replying with more detail)

    Thank you again for the reply, sorry I was still short cutting on details (honestly I didn't expect such helpful service) I'll check out the link when I'm next at my pc but I'm not sure it will help, I'll try to explain my situation  below.  

    So I am looking to run the ADS1294 in conjunction with an AFE4300 and the ADS1294 input will be on the same electrodes which will have a 50kHz+ signal. In my filtering I need to be careful to not ground that or the AFE4300 will not work properly.

    If I put something like a 10k ohm resister for the RC filter to block the AFE4300 signal from leaking to ground through the capacitor would I lose a lot of the ECG signal?ADCs and input impedance is a bit of a gray area for me.

    The other option is to filter above  the AFE4300  signal but I want to run that up to around 200kHz and suspect that will be hard to fit under the 256kHz nyquist with a cheap filter arrangement. I'll try out the program in the link for this though.

    The backup idea is something like a unity gain buffer or a bias tee but I'm not sure that is needed?

    Cheers,

    Cameron

  • 0 in reply to Cameron Elvey
    TI__Mastermind 38745 points
    Hey Cameron,

    Luckily the input to the ADS1294 is high impedance so I don't expect that to drain much current. However I can't count out the displacement current in the capacitor like you mentioned. That link I sent you is a simulation tool. I recommend you put the values you are considering in to a model and simulate to see how far from ideal it is with the RC filter. If you want me to take a look I can also try to simulate it and see what I find. Unfortunately I am not as familiar with the AFE4300 so it would take some time for me to get acquainted. However my sense is that if the filter resistors are significantly larger than the body impedance, it will not be much of an issue since you can think of it as a crude current divider.

    Regards,
    Brian Pisani