• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 449 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.

ADS1299-4: about precision, Gain Error

Hidetaka Kudo
Prodigy 50 points

Part Number: ADS1299-4
Other Parts Discussed in Thread: ADS1299

Tool/software:

Thank you for your continued support.
When we measured the waveform amplitude in our application by applying a 1.000Vpp sine wave between IN2P and IN2N, and between IN3P and IN3N, the amplitude was 0.986Vpp. In our example, this represents an error of approximately 1.4%. If you know the cause of this error, please advise.

For reference, we have attached the circuit diagram around the ADS1299-4.
Although it is designed for biological signal input, we are inputting non-biological signals for our application.
The ADS1299-4 registers are as follows.

- ID 3C
- CONFIG1 92
- CONFIG2 C2
- CONFIG3 E0
- LOFF 00
- CH1SET 60
- CH2SET 40
- CH3SET 40
- CH4SET 60
- CH5SET 00
- CH6SET 00
- CH7SET 00
- CH8SET 00
- BIAS_SENSP 00
- BIAS_SENSN 00
- LOFF_SENSP 00
- LOFF_SENSN 00
- LOFF_FLIP 00
- LOFF_STATP 00
- LOFF_STATN 00
- GPIO 0F

Note that The MCLK frequency input to the ADS1299-4 is set to 1.924608 MHz. VREF uses the ADS1299-4's internal reference voltage. When measured with a DMM (VOAC7412, Iwatsu), the VREFP pin was 2.0005V, the VREFN pin was -2.5000V, and the 4.5V VREF was accurate. Using the T3AFG200 (Lecroy) function generator, the amplitude of the 1Vpp sine wave measured with the DMM was 353.15mVAC (998.96mVpp).
The voltage between the terminals of the 100pF film capacitor C72, placed near the IN3P-IN3N pins of the ADS1299-4, was 352.33mVAC (996.54mVpp), indicating no significant attenuation. 

  • 0
    TI__Guru 69446 points

    Hello Hidetaka-san,

    Thank you for your post. 

    May I ask what is the frequency and common-mode voltage of the input sine wave? I see you are using ±2.5V analog supplies, so 0V common-mode voltage on INxP and INxN is ok.

    Have you taken into consideration the attenuation from the digital sinc3 filter?

    Regards,

    Ryan

  • 0 in reply to Ryan Andrews
    Prodigy 50 points

    Hello Ryan,

    Thank you for your prompt response.
    Upon reviewing Figures 27 and 28 in ADS1299 datasheet, I immediately understood the impact of the Sinc3 filter. While I had assumed zero attenuation, I now recognize that approximately 0.8% attenuation occurs at 50Hz and about 1.1% at 60Hz.

    The input sine wave frequency is 50Hz.
    I believe the common-mode voltage is 0V.


    Due to the nature of our application, we are sampling at 3759Hz with the ADS1299, but we are handling the signal at a downsampled rate of 1253Hz.
    We plotted the Sinc3 filter characteristics at 1253Hz. Since we are downsampling the data, we consider it to be roughly equivalent to equation (7).
    Since the input signal frequency is 50Hz, in our example, the gain is approximately 99.217% rather than 100%. Considering the Sinc3 filter attenuates the gain by about 0.783%, the true value is 992.17mVpp.

    Considering the accuracy of the function generator T3AFG200 is +/-1%, the accuracy of the digital multimeter VOAC7412 is +/-0.4% +15 digits, and the Gain Error of the ADS1299 is 0.1% (typical), 0.5% (max), we understand that the accuracy of our measured results is assured.

  • 0 in reply to Hidetaka Kudo
    Prodigy 50 points

    Hello Ryan,

    Since then, we also verified our application. The application uses an anti-aliasing filter (an 8th-order IIR filter with 4 blocks). For your reference, this filter is applied to data sampled at a data rate of 3759 Hz. It is subsequently downsampled to 1253 Hz. Since this filter was designed with a ripple bandwidth of 0.2dB, we found it could cause a maximum deviation of 2.2%.

    Therefore, we changed the ripple bandwidth to 0.01dB, reducing the deviation range to about 0.1%.
    Your insights in response to our inquiry have provided us with significant new understanding. We sincerely appreciate it.