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 Constant noise on all channels

Other Parts Discussed in Thread: ADS1299

Hello,

After performing many data captures from a custom designed PCB that entails the ADS1299 and a ATMEL 8bit micro-controller there appears to be some oddities in the signal.

PROBLEM: There is a constant ~13Hz signal that appears on all the channels when connected to the body.

I am using the ADS1299 with a unipolar supply (+5V) and with signal ended inputs (INxP) with a connection to a reference electrode through SRB1.

I have measured the analog supply pins and the VREF pins to see if I could locate the ~13Hz signal. It is not present at those pins and neither at the INxP inputs.

However, I have found that this 13Hz originates from the wireless transmission (2.4GHz Zigbee) emitting from the 8bit zigbee micro-controller, and this can be directly seen on the RFN/RFP 100ohm differential lines, the 13Hz corresponds to the request/transmission rate of the packets.

Measuring the internal noise of the system by shorting internally all channels using CHxSET [0:3] = 001 and sampling the data did not show any signals of ~13Hz.

I then shorted externally the active channels including the bias and references through 6.8k ohm resistors and sampled the data. The ~13Hz was present on only 1 of the channels. 

However, when placed on the body, this signal is present on all the channels. 

If the signal is present on all channels, should the BIAS be removing this? Below is a picture of the BIAS setup I am using. The BIAS_EIN is connected to the electrode front end shown below, and then to the body through an electrode.

This is the input filter and protection resistor, each channel has this on its input,

I have set the following registers,

PD_BIAS =1

BIASREF_INT = 1

PD_REFBUF = 1

BIAS_MEAS = 0

MISC1 [SRB1] = 1

BIAS_SENSP = [1 for all active channels, 0 for deactivated channels]

BIAS_SENSN = [0 for all channels] ** Is this correct if I am using single ended inputs only through INxP?

Does anyone have any thoughts or issues they see in my configuration?

  • Hey Mike,

    Your setup looks pretty good. One thing you could try is setting one of the BIAS_SENSN bits. If you think about it, you'd want as many inputs to the bias amp as there are measurement electrodes. Right now you have that except for one input corresponding to the reference electrode. This electrode is connected to every channel, but you only need to enable one of the sense inputs since there is only one electrode. That is to say you can configure BIAS_SENSN to 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, or 0x80. Let me know if that fixes the problem at all.

    One disclaimer: the bias amp cancellation will only work if the signals are truly "common" to all the inputs. If the signal is coupling on to the inputs with 180 degrees of phase between the positive and negative inputs i.e. they act like a differential antenna, then the bias amplifier will do no good.

    Regards,
    Brian Pisani
  • Hey Brian,

    I have tried setting BIAS_SENSN to 0x01, it did not seem to make a difference in the sampled output signal.

    Do you or anyone else have other suggestions?

  • Hey Mike,

    My guess is that the cables are acting like antennae. My recommendation would be to shield them as much as possible. Alternatively if it is still an issue, you could implement a digital filter to remove the noise. The filter would likely have to be pretty high order so I'm not sure if your micro could handle it. It sounds like you're sending the data wirelessly as soon as you get it, though. In that case you could offload the work of filtering to whatever host processor you are communicating with assuming it has more bus bandwidth than the microcontroller.

    Brian
  • Hey Brian,

    Thank you for your help. I reduced the output power of the RF TX and it seemed to have solved the constant freq signal present on the channels.