ADS1292 consulting

Hello,

Thanks for your post and welcome to our forum!

You're right that the ADS1292 was originally intended for ECG applications, but you can certainly use it for other applications as well. The answer to your question will depend on the noise performance that you requires to resolve EEG signals. Generally, EEG signals are smaller in amplitude, so the measurement circuitry requires lower noise performance.

The other features in the ADS1292, such as lead-off detection and the right-leg drive amplifier, are common to EEG applications as well.

Best Regards,

Thank you very much for your reply!I designed a EEG acquisition circuit, would you please help to see whether it can realize all the functions?

Hello,

If you would like us to review your ADS1292 schematic or register settings, we can certainly provide some feedback through the forums. If you have specific questions about using the features in the device, please post them in a separate thread.

Best Regards,

dbs2 .pdf

i have uploaded the  schematic ,and i use  Differential Input,before the  Differential Input is Band Pass Filter. and hope you give me some advice?thank you!

Hello,

Thanks for posting the schematic. Here are a few things I noticed:

1. DOUT and DIN are connected to the same net (SIMO0). Please separate.
2. The net BIAS should also be connected to RLDINV. R14 || C41 form the feedback path for the internal RLD amplifier.
3. P and N electrode are AC-coupled. You will need to shift the input common-mode voltage within the range described on page 22. Typically, the common-mode is set to mid-supply using the RLD amplifier output.
4. The OPA2735 does not appear to be a good fit for this application. Your current amplifier circuit shows the non-inverting inputs are connected to GND. However, the output can typically only swing to within 20mV of the rail, and simulation shows the output is clipping on one side. It will be necessary to shift the common-mode a bit higher, but the common-mode must remain lower than VDD - 1.5 V.

The simulation below shows the output signals for VIN = 10 mVp sine wave at 100 Hz. The clipping disappears if you increase the voltage at the non-inverting inputs by changing VCM to 500 mV. However, keep in mind that the differential offset voltage between the electrodes can reach as high as 300 mV. Therefore, you must increase the common-mode even further. 

At the same time, the input common-mode voltage to the ADS1292 PGA must be at least 1.1 V, assuming Gain = 6 and max input of 300 mV. Setting the VCM on the OPA2735 to 1.25 V seems to satisfy both conditions.

It would be easier for your design if you chose either a different amplifier or a larger supply voltage such that the design could support a mid-supply common-mode. In that case, the RLD output could be connected to the non-inverting inputs of the OPA2735 and you would not need to generate a separate 1.25-V VCM. I would recommend increasing AVDD to 3.3 V. The RLD output would be centered around mid-supply = 1.65 V and could be connected to the OPA2735 non-inverting inputs.

OPA2735_ADS1291.TSC

VIN = 10 mVp, VCM = GND, AVDD = 3 V

 

VIN = 300 mVp, VCM = 1.65 V, AVDD = 3.3 V

 

Best Regards,

Thanks !  This would takes me few days to understand. But i found if the band-pass filter use  bipolar supplies, and the fo of the band-pass filter approach to 20HZ,we can see that the input signal and the output signal is normal .  I have change the circuit , and can the circuit work properly?

Best wishes!

Thank you very much, your advice has given me great help.I have the last question，i'm not sure how N,P and Bias connect with brain . i think N and P is connected to the target section , and Bias is GND which connected far away from the target section .

best wishes！

Hello,

You are correct about electrodes P and N. These are used to connect your input signal to the EEG measurement circuitry. The BIAS electrode is placed remotely to set the common-mode voltage on the body to something that is controlled with respect to the measurement system GND - otherwise, the body is basically floating.

In your schematic, you are using a unipolar 3-V analog supply (AVDD = 3 V, AVSS = GND). Therefore, RLDOUT will be centered around mid-supply = 1.5 V. The RLDREF signal internal to the ADS1292 sets the common-mode voltage to mid-supply automatically when RESP2[1] = 1.

Best Regards,

RLDREF connect to AVDD if not used , so i connect RLDREF to AVDD . is this contradiction to "The RLDREF signal internal to the ADS1292 sets the common-mode voltage to mid-supply automatically when RESP2[1] = 1",or should RLDREF be set to floating？

Best  wishes and thanks！

Thanks,with your help,I will finish my design soon!

Best Wishes!