HiLetgoADS1256 5V 8 Channel 24 Bit ADC Data Acquisition Board Module High Precision AD Collecting Data Acquisition Card

Works well. I have tested all modes using the curiousscientist library available in Arduino IDE, paired with a Teensy 4.1 microcontroller. This is a complex device, with lots of settings and cross interactions between settings that requires studying the datasheet to understand. I shorted the inputs to ground and studied the volatage variation- which represents the noise which is internal to the device. Expect stability to about 100uV at highest sampling rate, and about 1uV at the lowest sampling rate. Your mileage may vary.

You get 8 channels of 24 bit single-ended or 4 channels of 23 bits differential with pre Amp gains of 1,2,4,8,16,32, and 64. The ADC Ref voltage is set to 2.5vdc on this board. Wiring to Arduino is easy. Software libraries and code are available for free online.

This board does not bring out the "Reset" line but I guess it usually isn't necessary anyway. I'm using the Curiousscientist library for ADS1256 with a Pi Pico microcontroller. Using max PGA gain (64x) and slowest sample rate (2.5 per second) I measure RMS noise with shorted input near 0.21 microvolts (in either differential or single-ended mode, and at either 0V or at 2.5V common-mode voltage). The ADS1256 datasheet predicts RMS noise should be 0.03 microvolts in these conditions, but that may require better shielding of the entire board from ambient RFI and air drafts (?) I am using the default series 1k resistors on each input line, but the thermal Johnson noise from them is still an order of magnitude smaller at around 2 Hz bandwidth, so it's not that.

I have tried 5 different ADS1256 boards with the Teensy 4.1, and these boards generally work the most reliably, even in "cycling" mode, after I made some changes to Curious Scientist's ADS1256 library. Some other boards also work, mostly, but can fail to work correctly particularly in "cycling" mode (usually after a short period of time). I also made some changes/additions to the library to avoid "cycling" mode entirely, and that works nearly as fast. I further tried 3 copies of yet another board that would work sometimes but mostly not at all. One can play around with the SPI speed and various time delays and that can help at times, but not reliably.