ADS125H02EVM: Testing lowest possible noise with EVM

Dave Young

Part Number: ADS125H02EVM
Other Parts Discussed in Thread: ADS125H02

Hello-

I've been testing the ADS125H02EVM and would like to get an idea of the lowest possible noise using the EVM as a baseline for my design. I took 30k readings and saw ENOB of 19.6 and noise-free bits of 16.4. Is this expected?, or is it possible to improve on it? Here is my setup:

Hardware setup:

EVM in stock condition with on-board +/-15V supplies enabled
Plugged into my laptop which is on battery (unplugged from the wall).

Set 0x03 MODE1 register to 0x21: Enable autozero
Set 0x06 REF register to 0x10: Enable internal reference, use int refP and int refN for the reference
Set 0x10 MODE4 register to 0x5B: Connect MUX as internal short to (HV_AVDD + HV_AVSS) / 2 and apply a gain of 128V/V

Here are my register settings and associated result:

over 2 years ago

+1 Bryan Lizon86 over 2 years ago

TI__Mastermind 35081 points

Hi Dave,

According to the ADS125H02 datasheet, the measured RMS noise at G=128, ODR = 20 SPS, and using the FIR filter is 74 nVRMS. You can input these values into equation 1 in the datasheet, along with the selected value of VREF = 2.5V, to calculate the estimated effective resolution.

So, effective resolution = 3.32 * log(2 * VREF / gain / noise) = 3.32 * log(2 * 2.5 / 128 / 74*10^-9) = 19 bits.

Since you got 19.6 bits in your measurements, I would say you are getting the lowest noise possible.

-Bryan

0 Dave Young over 2 years ago in reply to Bryan Lizon86

Intellectual 520 points

Awesome! Thanks so much for the sanity check and the excellent explanation with calculation. This is great.

Data converters

Data converters forum

ADS125H02EVM: Testing lowest possible noise with EVM