The resolution of a load cell is highly dependent on its capacity. That's because the higher the resolution and the higher the capacity, the smaller the voltage change. For example, many companies market a load cell with a capacity of 1 kg with a resolution of 0.1 grams, but there are no load cells on the market with a capacity of 100 kg with a resolution of 0.1 grams (if there is one, it would be very expensive).
Then, I found that measuring small voltages was difficult because of the influence of the thermoelectric EMF, like Seebeck effect. We can see on that website, the Seebeck Coefficient of Cu-Cu is ≤0.2 µV/°C. I think, the temperature difference in output circuit in the load cell is not much different, maybe less than 0.1°C because load cell use low current. So, for a temperature difference of 0.1°C, the noise voltage difference is roughly only about 0.02 µV. Then, temperature changes take time, so the thermoelectric EMF noise also takes time to appear.
So what makes it so difficult to build a load cell with a large capacity and high resolution?
