• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 46 subscribers
  • Views 1329 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.

ADS1298 ECG with fingers / dry contact

Govind M
Prodigy 115 points
Other Parts Discussed in Thread: ADS1298, OPA2333, INA333

Hello,

I am working on an application that needs to measure a persons ECG from his/her fingers (right and left thumb). Currently we are using a biopotential amplifier circuit using discrete components and the ICs INA333 and OPA2333. This gives a reasonable signal for many people but still is quite poor compared to leads connected to the chest. I am trying to evaluate a solution using ADS1298 series chips, to reduce the component count and hopefully improve the SNR. I have the ADS1298ECG-FE evaluation board with me.

I get very good results with the simulator but when I use finger leads, the signal is very noisy and has a terrible baseline drift. I have attached a sample waveform illustrating this along with the post. I have tried using the RLD derived as RA+LA internally and connected to the index finger. It does not seem to help much. At the input, the fingers are connected directly to the ADS1298 channels via a metal plate electrode and a wire. These are the same metal plates we use in the current application.

What could I do to improve the SNR and have less baseline drift? Should I have a pre-processing / anti-aliasing stage before the ADS1298? Any suggestions would be appreciated.

Thanks a lot!

Govind

  • 0
    TI__Mastermind 43915 points

    Hi Govind,

    Could you send you schematic of how you have your front end set up. You need to bias the body or the inputs somehow to keep the inputs within range helping with the baseline drift. One way you could do something like this is to AC couple the inputs using something like a 1uF capacitor and then on the ADC side of the capacitor, use weak pull up resistors to the RLD (1Mohm). Is this similar to how you have your design set up or are you doing something different?

    I have to state that the ADS1298ECG-FE board was not designed to be used directly with a human. There is no protection circuitry on the board which is required when used with a patient.

    Regards,

    Tony Calabria

  • 0 in reply to Tony Calabria
    Prodigy 115 points

    Hi Tony,

    Thanks for your reply, I was connecting the inputs directly to the ADS lines (on JP36). There was no specific front end circuit. In our discrete component based circuit, we had an integrator feedback at the instrumentation amplifier  output stage to help reduce the baseline drift. I tried your suggestion to add the capacitor and pull up the signal lines to the RLD output, and it improves the drift. However the SNR is almost the same as our discrete component circuit. Any ideas on how to improve this? Or is it just an unavoidable result of measuring from the thumbs instead of chest leads?

    I am just using the board to evaluate the performance of the chip, I won't be using it on any patients (other than myself and a few colleagues) :). By the way, what specific protection circuitry are you referring to?

    Govind

  • 0 in reply to Govind M
    TI__Mastermind 43915 points

    Hi Govind,

    Typical patient protection circuitry could include something like a two stage RC networks, neon bulbs, etc. I cannot legally comment and tell you how to design the patient protection due to liability reasons so you will have to research/talk with others on the forum to get help there. Defib protection may be required as well to protect the circuit from high charges during defib. A 1uF AC coupled input capacitor helps clean up the input here on my end but you need to also make sure you are biasing the inputs of the ADC within range. We have used a 1Mohm resistor to the RLD output to use the RLD to bias the inputs within range. This occurs right at the inputs on the "ADC side" of the AC coupling capacitor.

    It is hard for me to comment without seeing any pictures of your digitized signal. If you pass that along, I can look to see if it compares to what I have seen in the past using our simulator. I have seen very clean results using the thumbs of a digitized ECG signal from others on the forum.

    Regards,

    Tony Calabria

  • 0 in reply to Tony Calabria
    Prodigy 115 points

    Hi Tony,

    My application is more on home use and not clinical, so I don't need to worry about defib etc. Anyway I asked about the protection circuitry for information, that's all. Thanks for your reply and for offering to have a look at the signal. I have attached a sample recording using thumb leads. It's 10k samples at a rate of 500sps. if you plot the values in the CH2 column, you can see the signal.

    Let me know if you need more information.

    Thanks,

    Govind