ADS1256 reducing noise

Frank Schwartz

Part Number: ADS1256
Other Parts Discussed in Thread: REF5025, OPA350, REF6025, ADS1255

Hi!

I look all answers related to ADS1256 oise but still I do not find a solution for my problem. So, here I am asking on my way: is this level of noise acceptable for my configuration or not.

I think it is still too big....

1.) For the start I bought this ADS 1256 board: www.aliexpress.com/item/2046677415.html

I am using this module in combination with Arduino Mega2560.

After almost one week I managed to operate with this moduel and I am getting approximately correct results.

2.) After that I made my own dev board with ADS1256, to be sure that I know hot to manage with this chip. After few repairings, my module works,too.

But now, I am testing the performance of my board and the module I bought from internet anbd I found out, that according to my opinion, the nosie level is too big,

at the bought module and on my own.... The results are very similar:

PGA=1, buffer =Off, sampling speed I changed: 500SPS, 1000SPS and 30000SPS. I am using only signal from 0 to 5V, single ended measurement. The 5V signal I generate with precise 5V calibrator, battery power supplied. AINCOM I have connected to the AGND. Analog and digital ground are separated, connected together only in one small point under the ADS1256 chip...

Here are the results @5V DC input signal. I memorized the minimum and the maximum value:

5V @500spsp:

My module: MIn=7827738, MAX=7831099, difference=3361

Bought module: MIn=7819528, MAX=7822641, difference=3113

5V @1000spsp:

My module: MIn=7826321, MAX=7830670, difference=4349

Bought module: MIn=7816679, MAX=7820689, difference=4010

5V@30000sps

My module: MIn=8343959, MAX=8352881, difference=8922

Bought module: MIn=8334841, MAX=8344354, difference=9513

So the results are logical, less sps I have , the difference etween the min and max value at the same voltage is smaller. But I think this difference is still too big....

For example the difference @500sps is app 3000, comparing to full 23 bit value=8388607 this represents 0.035%.

Is this according to expected? I expect less than 0.01% oe am I wrong? Did I miss something?

Thank you for your help and best regards!

over 3 years ago

0 Bryan Lizon86 over 3 years ago

TI__Mastermind 34921 points

Hi Frank,

Have you tried shorting the ADC inputs together and take a reading to measure the noise of your system? You might try this on both boards just to set a baseline noise figure.

What reference voltage are you using? A precision VREF such as REF5025? If your reference voltage is not clean and stable you could very well see issues like this. The fact that your code ranges are so different for the different sample rates suggests something might be going on with the VREF. You should be getting close to the same result each time if you are sampling the same signal. If you have a schematic to share that would help so I can see what your signal conditioning circuitry looks like.

We also generally recommend against split ground planes, instead suggesting one solid ground plane. I am not sure this is the issue you are seeing, but just something else to consider.

-Bryan

0 Frank Schwartz over 3 years ago in reply to Bryan Lizon86

Data converters

Data converters forum

ADS1256 reducing noise