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.

ADS1294R: ECG Front-end

Part Number: ADS1294R
Other Parts Discussed in Thread: ADS1298R

Hello,

I need some technical support about designing the ECG sensor with using front-end ADS1294R.

I am looking for typical Application for three channels ECG and respiration measurement using internally-generated modulation signal on the same electrodes used for ECG to have an acceptable signal to noise ratio.

Also, according to the “table 3. Input-Referred Noise in High-Resolution Mode” with output data rate of 1024Hz, the noise level should be around 3.9μVPP.   With using register “CHnSET, (address = 05h to 0Ch)- Bit (2:0)- Input shorted”, We measured around 3 μVPP, however on human body the noise level is around 30 μVPP. Do you have any recomendation or a specific application note to reduce the noise level?

 

I am looking forward to hearing back from you.

Best regards,

Rasool

  • Hi,

    May I ask where do you see " with output data rate of 1024Hz, the noise level should be around 3.9μVPP. " ?  I don't see such entry in the Tables.

    What device and how did you measure the noise on human body? 

    Do you happen to have/use any ADS1298R evaluation kit/board(EVM)?

    When you say "typical Application for three channels ECG and respiration ", do you mean just Right Arm(RA), Left Arm(LA) and Left Leg(LL)? Are you using or will you plan to use the Right Leg Drive(RLD) amplifier feature? RLD could help reduce common node noises.

     

    It definitely makes sense to have higher noises when the IC starts interacting with external environment as there are many other factors, components and variations along the signals chain/path contributing to the noises.

    There are some steps and guidelines people could do to improve the noises, some are listed here -

    0. Did you try to use any ECG simulator acquire signals before interacting with human body? It's suggested to test with ECG simulator to understand the device and system first.

    1. If not using EVM, do you follow the ADS1298R datasheet Section 12 Layout guidelines to design your PCBA? Any noises aggressors near the signal lines/chain/path? ADS1298R EVM user guide provides schematics, which might be a good reference to start with.

    2. If not using EVM, how clean are your power, voltage sources, reference voltage sources along the path?

    3. Figure/find out the noise sources and frequencies. If possible, calculate the Signal-to-Noise ratio(SNR), and see whether such SNR meets the products' needs or applications? 

    4. Try to see if one could improve/reduce the noises from the sources or along the path that it may cause noises coupling. e.g. using noise decoupling capacitors, analog low pass and/or high pass filters and/or digital low pass, digital high pass filter, etc.  ADS1298R EVM user guide provides schematics, which might be a good reference to start with.

    5. Are you using or will you plan to use the Right Leg Drive(RLD) amplifier feature? RLD could help reduce common node noises.

     

    Thanks

  • Dear Mr. Yang,


    Thanks for your response.
    Although, we have EVM board but we are using our own device based on ADS1294R.The device has 4 electrodes RA, LA, LL, RL and can measured 3 channels ECG signal and respiration. The front-end is configured as High-Resolution Mode, 5V analog supply and 4V reference. RL electrode is connected to GND by a series resistor of 110K (we’ve got no better result by using RLD amplifier). The PCB is isolated by galvanic isolation method. The device has a microcontroller and sends the information via UART communication to Matlab or our main device which is powered by a battery.


    In datasheet page 22, section 8.1-Noise Measurements, table 3, It is written for output-data-rate 1000SPS Input-referred noise should be 3.9μVPP. I set the register “CHnSET” to have Input shorted for measuring the noise. I recorded the peak-to-peak noise level around 3.3 μVPP. Furthermore, I made a short circuit directly on electrode (cables length around 30cm) and I measured around 4.3 μVPP as noise level. But as expected on human body it is different.

    Regarding ECG signal simulator, we are using FLUKE ProSim. The signal quality looks very well. I have attached a picture of a sample measurement with ECG-simulator.
    The sensor is powered by our main device with a battery as a power source and the ECG sensor is galvanic separated and also has some regulators to provide stable voltage sources.


    RLD amplifier feature did not improved the results and signal to noise ratio is differ from person to person. But channel one has lower signal to noise ratio. I also attached a picture of human body signal.by the way if I figure it out it will be better to use RLD I would definitely switch on this feature.


    Because of measuring respiration and ECG on same electrodes, I cannot use a low pass filter with low cutoff frequency. Respiration is configured with 32KHz modulation frequency and having a big capacitance reduces the amplitude of modulation signal. Now I am using 1K series resistor and 100pF just for EMC purpose and as an ECG input chain.


    I am looking for an application diagram with appropriate components acording to my application.

    Thanks for your time.


    Best regards,
    Rasool

  • Hi Rasool,

    Appreciate for your detailed description and explanation.

    May I suggest some more steps?

    1. Could you use your system to configure the ADS1294R in your system to do the internal test signal test? And to see if the noises/interference still there or not?

    2. TI has ADS1298R EVM instead of ADS1294R, but it's just number of channels difference. If you have ADS1298R EVM, could you replicate your test setup using ADS1298R EVM and use the GUI to see if the noises/interferences still there or not?

    3. Regarding to the circuit reference design for ADS12984R, please refer to ADS1294R datasheet and ADS1298R EVM user guide page 62 to 66.

    4. What might be the interference/noise frequency? Is it 50/60Hz?

    5. Note - If the RLD function is not used, power down the amplifier using the PD_RLD bit (see the CONFIG3:Configuration Register 3 (address = 03h) (reset = 40h) section for details).

    Please let me know whether you would like to bring this public discussion to offline email discussion, so that you could share portion of schematic with me and we can do some reviews?

    Thanks