Thanks for your reply. So the SOP8 chip is LM75 not an EEPROM, there is internal EEPROM in ATmega328.
Yup. Places where penguins nest tend to have extreme temperatures, so I thought it would be useful to include the LM75. I don't think I ended up using it for compensation, and it isn't logged at the penguin weighbridges. I think I do log it at another (wallaby) weighbridge installation in Queensland.
The only need for the EEPROM is calibration constants and basic settings, which change pretty rarely. Possibly once a week at most, if the researchers follow procedure and there is any noticable drift.
I still feel like that the use of feedthrough capacitors is a little overkill here, they are expensive.
Sure, but I wasn't building these in any significant volume, and I believe I already had the same feedthrough caps left over from another design.
This was actually an urgent redesign to replace failed boards using a 20 y.o. design where the parts were no longer available. I wanted something that worked as well as possible first go!
I developed two firmwares - one done the way I would do it, and one that was backwards compatible with the historical protocol in use and its idiosyncrasies
When I did noise tests, the 24 bit ADC gave me 19 effective bits which I was perfectly happy with. Better than the original design, and we only needed 1g resolution.
BTW, I'm from China, we don't even use ADS1232 anymore, we use HX710/711, very very cheap.
I've used the HX711, and contributed to the Arduino library at the time. As simple as it was, I didn't like it one bit.
Actually, V2 of the above design used the HX711. I assembled it quickly over a weekend as a stop-gap. It gave occasional whacko data which caused some grief, and required filtering. Could have been an implentation issue, but I wanted something better documented and that I felt I (and others) could trust. For something being used in scientific research, I wasn't going to risk using a poorly specified part.
Back on topic...These have been in the field for about 3 years, suffering widely varying temperatures (not Antartic conditions, but unprotected coastal Australia), as well as costal salt air, and even full submersion after heavy rain. Running constantly with no reported problems. Knock wood...