• State TI Thinks Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 4 answers
  • Subscribers 48 subscribers
  • Views 3392 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: ADS1299EEG-FE 10 channel bias and reference connection

Ali Eissa63
Prodigy 70 points
Part Number: ADS1299

I'm using the ADS1299EEG-FE to carry out a test on an 8 channel EEG system. The total number of electrodes is 10: 8 channels + 1 reference + 1 patient bias electrode. 

I am connecting the 8 channels to the corresponding pins on J6 [36, 32, 28, 24, 20, 16, 12, 8], and I was under the impression that I could connect the reference electrode to pin [6] on JP 25, apply a jumper to pin 1,2 on J6 to route the reference signal to all the negative inputs through SRB1. I also wanted to connect the patient bias electrode such that it would drive the patient to mid supply voltage, and I am doing so by connected the reference electrode to pin [4] on JP25. 

I havn't tested the system yet, because I am not sure about any of the reference or the bias electrode connections, and I have tried reading through the user's guide, but I couldn't clearly understand how the connections should be made. Most of the connections I have currently made are from suggestions on the forum from similar set ups.

  • 0
    TI__Guru 66196 points

    Hello Ali,

    Welcome to our forum and thank you for your post!

    How you configure electrode measurements in your appliacation is up to you, but I can try to provide some clarity on the EVM schematic in the User Guide.

    For starters, BIAS_ELEC can either be driven by an on-board mid-supply voltage, created by a resistor divider circuit and a buffer, or by the internal BIAS amplifier in the ADS1299. Simply select which one you want to use using the appropriate jumpers. If you're using the internal BIAS amplifier, you will also need to set the appropriate bits in the CONFIG3 register. Please note that, given that this is an evaluation platform, we have not designed or certified the proper protection circuitry required for such applications.

    It may be easier to connect the BIAS electrode on JP25[2,4]:

    To connect the reference electrode to the negative inputs using SRB1, tie the reference electode connector to JP25[6] and install JP8. This feature also requires an additional register setting (MISC1[5]).

    Hope this helps!

    Best Regards,

  • 0 in reply to Ryan Andrews
    Prodigy 70 points

    Hello Ryan,

    Thank you for your assistance, this was certainly helpful. 

    As for the Bias electrode, I'm connecting it to JP25[4], I was wondering if I have to change anything in the bias control registers from their default state, since it seems to still not work.

    CONFIG3 has the [Reference Buffer] and [Bias Measurement] open, [BIASREF signal source] fed externally and the [Bias Buffer] is disabled. 

  • 0 in reply to Ali Eissa63
    TI__Guru 66196 points
    Hi Ali,

    You will need to enable the BIAS buffer and select the internal BIASREF voltage in order to drive a mid-supply common-mode on the bias electrode. For the ADC to make conversions, the internal reference buffer also need to be enabled.

    Best Regards,
  • 0 in reply to Ryan Andrews
    Prodigy 70 points

    Forgive me for repeating my question. I have attempted to do as you suggested, but the its still not working. I have attached a picture of the setup.

    The orange wire connection to JP25[2] is the bias electrode.

    The green wire connection to JP25[6] is the reference electrode. 

    The rest are the 8 channels, and the jumper settings are as you see.

    What am I messing up, because I can't seem to see it.

  • 0 in reply to Ali Eissa63
    TI__Guru 66196 points
    Hello Ali,

    Why do you say that it is not working? What does your input signal look like and what results are you getting?

    It looks like you elected to use the on-board, buffered mid-supply voltage by installing JP6[1-2]. Is that what you meant to do? In this case, you do not need to enable the internal BIAS amplifier inside the ADS1299. The rest of your electrode connections look correct.

    We really need to see the input signal results and corresponding register settings in order to provide much more guidance. Also, given that this is for EEG, it will be easier if you could use a more reliable signal source as you get the system set up, such as a function generator with a sine wave output.


    Best Regards,
    Ryan
  • 0 in reply to Ryan Andrews
    Prodigy 70 points

    I  apologize for the late reply, the labs were closed through new year. 

    The setup currently is connected with EEG electrodes to an apple to get some noise. What I intended for the bias electrode to do was to use the common mode rejection to reduce the 50 Hz noise. 

    I've tried both JP6[1-2] and also connecting JP1[1-2], but the 50Hz value is not attenuated by the bias circuit at all, such that I get the same 50Hz amplitude with or without the bias electrode. 

    The other part is that I am not getting mid supply at R8, rather random different values on different setups. I'm currently reading 1V at one end and 0V at the other end of the resistor with JP1[1-2] installed, and with the corresponding register values: 

  • 0 in reply to Ali Eissa63
    TI__Guru 66196 points

    Hi Ali,

    Happy New Year!

    JP6 can only be used to provide a DC voltage to the BIAS electrode. There are no electrode signals included in the feedback path, which are needed to sense the common-mode voltage on the electrode inputs and derive the appropriate cancellation signal.

    The BIAS amplifier will sense the individual electrode inputs as you set the corresponding BIAS_SENSP/N bits to 1. Right now you only have IN2N selected. Is this one of your electrode inputs? You don't have to select every electrode that you use, but you may experiement with selecting more than one.

    How are you measuring the 50-Hz noise? Are you collecting and analyzing data or just probing the inputs? Scope probes can introduce a lot of extra noise as well, so it's best to just short the electrodes together (inputs and BIAS) and look at the noise in an FFT or histogram. Also, remember that you won't completely eliminate this noise, even with proper common-mode cancellation.

    I would start with all BIAS_SENSx bits set to 0 and fix the mid-supply voltage problem first. Remember that you are using bipolar supplies (+/-2.5 V), so mid-supply will be close to GND (0 V). One side of R8 is connected to the BIAS amplifier inverting input (BIASINV). This node is a virtual short to the amplifier's non-inverting input, which is conneted to the internal mid-supply BIASREF voltage.

    Check out this application note:

    Improving CMRR with RLD:

    Best Regards,

  • 0 in reply to Ryan Andrews
    Prodigy 70 points

    Hey Ryan, 

    Thank you a lot, that really helped. 

    I eventually realized that the current Rf and Cf values for setting the bandwidth of the RLD loop is set at 40hz, according to the equation fBW = 1/(2 pi Rf Cf).


    Since my setup is in congruence with what you suggested, I believe I will need to adjust the resistance and capacitor values to actually include the 50hz value that I am working with.

    I hope that this is the correct thing to do, I will update you if that turns out to be the solution. 

    Also please let me know if I misunderstood the schematics/equations from the manual and the RLD file, or if my decision to change the resistance and capacitance values will not have the desired effect.

  • 0 in reply to Ryan Andrews
    Prodigy 70 points

    Update:

    I have changed the capacitor C20, and used a 4.7pf capacitor instead to increase the bandwidth of the bias amplifier.

    This seemed to have no effect on the output. I'm merely connecting my jumpers to an apple to see how the 50hz noise is affected by the RLD circuit.

    I have attempted different feedback electrode selection, but the amplitude of the 50hz and its multiples is not affected.

    Output FFT: 

    Config3:

    Registers:

  • 0 in reply to Ali Eissa63
    TI__Guru 66196 points
    Hi Ali,

    Can you check the DC voltage of the BIAS amplifier and make sure it is close to mid-supply? With all 16 inputs selected for common-mode derivation, the gain has been set really high (392k / (16 220k's in parallel)) = 28.509 V/V. This is probably saturating the amplifier output. It's fine that you increased the closed-loop bandwidth beyond 50 Hz, but remember that the cutoff frequency is where the magnitude drops -3 dB from the DC gain. Therefore, the response may still be high enough above unity gain (0 dB) at that frequency to be effective.

    Regards,