ADS1256: Increasing accuracy of ADS1256 reading

Hello Bryan,

Thanks for the reply, actually i want to read the pressure sensor data with ads1256 and now i can able to read the ADC value but i want to achieve precision in adc reading because i am dealing with ads1256 with nano volts changes.

for example,

1. Now i want to fetch the micro/nano voltage changes  ads1256 is it possible .?? because if i to convert the voltage into Pascal or PSI unit means i need to show even small changes of input pressure its like the pressure value is 3.5786839 PSI or  24674.15691887 Pascal so here once i stop input pressure means value starts decreasing very little because small leakage in pressure so at this point pascal value fluctuating lot like after decimal point the first digit start fluctuating.

so my point is i couldn't control the pressure value here because in software its just calculation and conversion if any small changes in ADC output then it affects large changes in Pascal because we are multiplying with a large value.( 1 PSI = 6894.76 PASCAL)

Finally, i am thinking if I can able to read the very minute changes like extremely small changes in adc reading means then can show the variation of pascal value in the 4th or 5th digit of the decimal point so i hope you understand my problem.

Already i have tried many ways to increase the accuracy like taking the average value and once the device power-up value starts from zero like subtracting with offset value but i couldn't achieve anything.

please help me to resolve this problem.

thank you.

Hi sagar m,

Some questions for you:

1. What is the input voltage range you are using? In other words, what is the output voltage range of your sensor?
2. What is the reference voltage you are using?
3. What data rate are you using?
4. What is the resolution you are trying to achieve relative to the sensor output voltage? For example, "My sensor has an output voltage range of 0-5V and I am trying to achieve 1uV resolution / accuracy"
5. Do you have the ADS1256 buffer enabled or disabled?
6. Have you tried using the ACAL feature on the ADS1256? If so, what was the result?
7. Can you please send us a schematic to review?

-Bryan

Hello Bryan,

Thanks for the reply, 

1. My sensor output voltage range (is 0-103mV) so that's why i choose to gain 32 

2. reference voltage 2.5 V

3. datarate is 10SPS

4. resolution is 24-bit and i want it in nano volt or pico volt like i wanted to read extremely small voltage changes in the input because its very useful to show the pressure value variation when i convert it.

5. Yes i enabled the buffer by writing 0x02 to the status regster.

6. No I didn't try please explain more regarding this.

Please let me know if anything required.

thank you.

Hi sagar m,

Some comments about your schematic:

• The reference voltage appears to be generated by a switching regulator and a resistor divider. Why are you not using a precision voltage reference for this application when noise and DC performance are so critical?
• You show that the INA has a gain of 100. Are you also using a gain of 32 in the ADC? If so, what is the actual output signal from the sensor (it cannot be 100mV if you are gaining it up by 100 then by 32)?
• The noise of your system will be dominated by the front-end gain stage because it has a large gain (100). According to the AD620 datasheet, the 10 Hz noise at G=100 is 280nVPP typical. This will effectively be your noise floor, assuming no other noise in the system. So this is the smallest signal you can measure with this system
• What is the point of the OP07 in your system?
• I would include LDOs on all supplies to ensure that you don't see power supply noise in your system
• Do you have separate ground planes on your board? I see a ferrite bead connecting your two ground symbols. Split ground planes can lead to all sorts of issues. You can reference section 12 in the Precision Labs content for more information: https://training.ti.com/ti-precision-labs-adcs

-Bryan

Hello Bryan,

2. here for 2nd point you said INA has gain 100 but i didn't use op-amp and buffer here due some strange value so i removed those two.

actually, i couldn't validate the op-amp itself please help me to validate the op-amp and if i used op-amp with gain 100 then in software what gain should be  use.??

3. Please suggest me comfortable op-amp which can suitable for my application.

4. op07 acts as a buffer.

thank you.

hello bryan,

I wanted to achieve more accuracy and without any fluctuation of adc data so for that, i am planning to use ADS1261 with your full suggestions.

1. Please tell me that i can use ADS1261 adc to read pressure sensor data with more accuracy and no fluctuation.??

2. Is there any dev kit available to for ads1261.??

3. If dev kit is available means please let me know how can i achieve more accuracy and no fluctuation.

i need your full support for ADS1261 adc reading and achieving fewer noise data read.

please let me know your feedback.

thank you.

Hello Bryan,

thanks for the reply, Ok please have look at the video for what  exactly i need,

in this video, i am not inputting any pressure so it's just like an idle condition so when there is no pressure you can observe i got 9 .4 mV and in the video, its in nano volts and the total psi value is 0.044 PSI but the value varies or fluctuates after 3rd / 4th  digit in the same way i need pascal to vary for that i need to fetch extremely small changes in ADC is that correct.??

so for that, i will try to test with the ADS1261 dev kit, and please let me know if any other changes are required in my design to suppress the noise.

thank you. 

Hello Bryan,

thanks for the reply, 

Bryan Lizon86 said:

which I assume means nanovolts - is that correct?

Yes,

Bryan Lizon86 said:

So you are measuring 0.00000095 nV? But in your last post you said you were measuring 9.4mV, so now I am confused.

No sorry,its actually  0.94mV the same value in the video so i have converted it into nano volt.

Bryan Lizon86 said:

What voltage are those values e.g. "0.00000095" supposed to represent?

its differential sensor output which is fed into ADS1256 and ADC output is represented as nano volts. is that clear.??

Bryan Lizon86 said:

Have you calibrated the sensor? Ideally with no load applied the output should be 0V.

i am calibrating the pressure value to convert the Nano voltage value into PSI and not calibrating the ADC output value here its just normal voltage conversion from the raw value of ADC  but the ADC value should be zero as you said but i got some values eventhere is no load i mean there is no pressure.

that's why in PSI and Pascal get varies even there is no pressure.

i have tried lot to make it zero but still some fluctuations in ADC reading like in Nano Volts, so is there any solution for this.??

thank you.

Hello Bryan,

I am little bit confused like i  have gone through the document i got some info like 

is that need to set up circuit and connect those four wires of sensor to ADC .??

Can you please explain it deeply?

thank you.

Ok Bryan thanks for the info i will read the document and let me know the update.

Hello Bryan,

I have gone through the document, and there is a wheat stone resistance bridge used with 4 wires like for my application, and then positive  and negative out fed to ADC AINP and AINN but i have question here 

for example  i have differential pressure sensor with pins 

1. GND

2. sensor o/p  NEGATIVE 

3. 5V

4 sensor o/p POSITIVE 

so as off now i fed these two sensor output pins to my AD620A op-amp as said in the previous thread  

and the it leads to ADC.

and  DP sensor looks like this with 4 pins

 I  think the output comes from the sensor its self-differential voltage which is fed to op-amp and then it leads to ADC. so here how can i use the wheat stone bridge to connect the sensor or how can i calibrate the device i am confused.

please tell me the way of calibration.

thank you.

Bryan Lizon86 said:

, you will not need to do this if you just apply the sensor output directly to the ADC inputs.

For the sake of achieving higher voltage calibration, i have used the AD620A, like gain 32 is 156.25mV and i need to calibrate it to 103mV so that's why I have used AD620A. 

is that the same thing you said.??

Bryan Lizon86 said:

This provides step by step instructions for calibrating a resistive bridge sensor. As I stated previously,

please don't mind i have silly doubt, like the wheat stone bridge is inside the sensor.?? or i need to make a bench setup of wheat stone bridge.??