• State TI Thinks Resolved
  • Locked Locked
  • Replies 15 replies
  • Answers 2 answers
  • Subscribers 47 subscribers
  • Views 2148 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 - precordial signal quality issue

Cesar Ulises Gutierrez Ramirez
Prodigy 210 points
Other Parts Discussed in Thread: ADS1298, ADS1298ECGFE-PDK, TINA-TIPart Number: ADS1298

Application: ECG device

Question type: Development and troubleshooting

Support topic: /Product/Device selection assistance/

Description of the problem: Hi everybody,

I´m using a ADS1298 for getting ECG signals.
Channels 2 and 3 for bipolar signals are OK, but the rest of the channels present noise when reading the signals, these channels are used for precordial signals.

I setted the configuration of registers showed on the TI software ADS129xECG-FE for each channel, but the problem remains unsolved.

Please, any information will be appretiated.




  • 0
    TI__Mastermind 49906 points
    Hi Cesar,

    I am moving your post from Medical Forum to Precision Data Converter forum where the ADS1298 device is supported.
  • 0
    TI__Mastermind 38745 points
    Cesar,

    Can you provide your schematic and some data showing the issue?

    Brian
  • 0 in reply to Brian Pisani
    Prodigy 210 points

    Hi Brian

    I attached information of the problem we have with the ADS1298. For the schematic we made a copy from the circuit of the development board (ADS1298ECGFE-PDK).

    Also, when working with TI board, we had the same problems about noise with precordial leads. That is, we connect the ADS1298 board to our mother microprocessor board (TWR-K53N512) from freescale, via SPI and we have the same results with precorials (V1-V6).

    The settings we are using in our code for the registers of the ADS are the followings:

    0x86, /*CONFIG1*/

    0x10,  /*CONFIG2*/

    0xDC,  /*CONFIG3*/

    0x03,  /*LOFF*/

    0x00,  /*CH1SET*/

    0x00,  /*CH2SET*/

    0x00,  /*CH3SET*/

    0x00,  /*CH4SET*/

    0x00,  /*CH5SET*/

    0x00,  /*CH6SET*/

    0x00,  /*CH7SET*/

    0x00,  /*CH8SET*/

    0x00,  /*RLD_SENSP*/

    0x00,  /*RLD_SENSN*/

    0xFF,  /*LOFF_SENSP*/

    0x02,  /*LOFF_SENSN*/

    0x00,  /*LOFF_FLIP*/

    0x00,  /*LOFF_STATP*/

    0x00,  /*LOFF_STATN*/

    0x00,  /*GPIO*/

    0x00,  /*PACE*/

    0xF0   /*RESP*/

    Also we attached images of some obtained signals, showing the problem of the precordial signal.

    Please, if you need more details, let us know.

    Greetings!

    ADS schematic.pdf

  • 0 in reply to Cesar Ulises Gutierrez Ramirez
    TI__Mastermind 38745 points
    Hey Cesar,

    The important registers to look at when it comes to the chest leads are the WCT1 and WCT2 registers. Can you also send me the values in those registers?

    Regards,
    Brian Pisani
  • 0 in reply to Brian Pisani
    Prodigy 210 points

    Hi Brian,

    Thank you for your support. The values are as follows:

    WCT1 = 0xc8 

    WCT2 = 0x0e

    Please let me know your comments.

    Best regards.

    César

  • 0 in reply to Cesar Ulises Gutierrez Ramirez
    TI__Mastermind 38745 points
    Cesar,

    The WCT amplifiers are disabled. This means the INxN inputs to all the chest lead channels will be floating. This is why you see such strange results. Change the registers to these values and I think you will see the results you are looking for:

    WCT1: 0x0B
    WCT2: 0xD4

    Regards,
    Brian Pisani
  • 0 in reply to Brian Pisani
    Prodigy 210 points

    Hi Brian,

    Thanks for the information. We already did the changes, but we ara having another problem, drdy pin seems not to be active, so the ads is not sampling.

    Any idea if the change of the registers affect in this problem?

    Best regards.

    Cesar 

  • 0 in reply to Cesar Ulises Gutierrez Ramirez
    TI__Mastermind 38745 points
    Hey Cesar,

    Unless the device was accidentally sent the STOP command, I doubt that the register change would have caused that. If you change the registers back to what you had before does the problem go away?

    Brian
  • 0 in reply to Brian Pisani
    Prodigy 210 points

    Hi Brian,

    The problem of the sampling was on our code, we already solve that. But with the changes of the WCT1 and WCT2 registers we have the same problem, the signal still noisy

    Best regards,

    Cesar.

  • 0 in reply to Cesar Ulises Gutierrez Ramirez
    TI__Mastermind 38745 points
    Hello Cesar,

    Have you verified the settings of those registers with a read? If you look closely at the settings I gave you, do they correspond to the correct settings for the WCT signal you wish to produce?

    Regards,
    Brian Pisani
  • 0 in reply to Brian Pisani
    Prodigy 210 points

    Hi Brian,

    We made some tests with the TI software of the ADS1298ECG-FE and analyzed the register values of all registers.

    We wrote the values in our device but the results are still the same, bipolar signals OK and chest signals with noise. Even with the values that you gave me (WCT1=0x0B, WCT2=0xD4) we have the same results. Attached you will find a document with the registers values and a figure of the signals.

    Please, let us know if there is any other configuration that we can use.

    Best regards,

    CésarRegister map and figure.docx

  • 0 in reply to Cesar Ulises Gutierrez Ramirez
    TI__Mastermind 38745 points
    César,

    In the latest figure, V1 looks much better than in your previous posts. It looks to me like a very small ECG waveform with a lot of mains frequency interference. If you were to filter the data with a filter that notches 60 Hz, does the signal look better?

    Brian
  • 0 in reply to Brian Pisani
    Prodigy 210 points

    Hi Brian,

    The latest figures were already filtered. In these new ones, we obtained all the 6 precordial, D1 and AVF signals.

    In the figures on Part I (attached below) are the signals obtained directly from de corresponding channels, and we noticed that the only ones that had noise are channels 8 and 1 (V1 and V6).

    On the figures on Part II we exchanged those channels for ch5 and ch6 to check if there were some changes on V1 and V6. With this exchange, the precordial signals didn´t have noise, but V3 and V4 (signals obtained from ch8 and ch1) had noise.

    We already did this test with other leads, and the problem exists too. Please, can you tell us if there is any procedure to take in order to improve signal quality in these channels (1 and 8); because as we said, the problem persist for these channels.

    Best regards,

    César.

    All_signals.docx

  • 0 in reply to Cesar Ulises Gutierrez Ramirez
    TI__Mastermind 38745 points

    César,

    What you're seeing is 60 Hz utility interference produced by mismatch in the transfer functions of the respective negative and positive inputs' RC filters.

    Note how the electrode inputs have the second order RC filter with a cutoff frequency of ~227 kHz. The WCT output that gets fed to the negative channel input has an RC cuttoff of approximately 5 kHz. This is due to the low pass transfer function of the WCT circuit. Look at page 36 of the datasheet. There are 30 kOhm resistors at the output of each WCT amp. Then, in your circuit, there is a 1nF capacitor at the WCT pin.

    The result of this is that the 60 Hz mains interference (which is common to all electrodes) gets filtered more heavily buy the WCT circuit and so it appears smaller at the IN1N input than it does at the IN1P input. This allows it to pass to the ADC. I simulated the ideal circuit in TINA-TI. Below I show the schematic where a common 60 Hz signal is fed to a single positive input through the RC filter and then to 3 other inputs that correspond to the limb leads. They are then buffered like they are on the WCT amplifier and sent to the negative input of a voltmeter. I attached the simulation file to the end of this message in case you want to simulate the behavior yourself or see how modifications will affect performance.

    The results illustrate my point about common mode rejection.

    If you want to fix this, I recommend setting the cutoff frequency of the WCT circuit much higher in frequency and also bringing down the cutoff frequency of the electrode RC filter slightly. Please download TINA-TI and play with the attached simulation file to find a good combination.

    Regards,

    Brian Pisani

    WCT_CMRR.TSC