I am trying to build a circular weight scale that relies on load cells to calculate the weight. I currently have the 4 load cells arranged in a cross under a circular metal sheet. Essentially, I have a load cell in each cardinal direction. Each load cell is connected to a channel on the PhidgetBridge 4-Input (SKU: 1046_0). Once I calibrate the weight scale, and try to put a weight in the North and South direction, the readings are very off with North bias reading a heavier mass and South bias reading a lighter mass. The East and West direction read fine when I bias any weights there. These load cells are connected to Channel 0 and 2 respectively.
To debug this issue, I have tried to rewire the 4-Input bridge but I consistently see that load cells connected to channel 0 and channel 2 give inaccurate readings. I also also tried a different 4-Input bridge. Does anyone have any solutions to why this might be happening?