Dear Support Team,
to get familiar with the capacitive liquid level sensing technology I build a test setup (picture below) with a FDC1004EVM using the Out-of-Phase technology. The dimensions of the setup are quite similar to the TIDA-00317 reference design. The copper tapes are 15mm wide, 0,036mm thick. The shields are 100mm, the references 20mm and the level 70mm long. The measuring gap is 4mm wide. As the insulator between the copper I used Tesa tape. The connections to the FDC1004 are (unfortunately unshielded) 0,2mm copper wires.
The whole setup does work quite reasonable. The submersed liquid reference is stable around 10pF, the environment reference about 7pF and both are suprisingly robust to external disturbances (hand in the vicinity or touch). The only problem is the level on CIN1, because it reaches saturation at 15pF already at a liquid level just about 10mm above the upper end of the liquid reference.
When the system is switched on, the two references directly give a stable value whilst the level measurement slowly (about 2 seconds) goes up to saturation. The level measurement is not totally out of range, because I can get the value down to about 13pF when tilting the bottle and by this lowering the liquid level to the lower end of the copper strips.
As the OoP technology depends on symmetry of the setup, I guess that the accuracy of the test setup is not good enough. I further guess, that I do need to achieve a capacity-wise symmetry, not only a mechanical one. The test setup has manufacturing tolerances of the copper strips about roughly +/- 0,3mm length, about +/-0,5mm gap width and of course a non-symmetric routing of the (unfortunately unshielded) connecting wires.
As the CAPDAC won´t work with the differential measurement of the OoP technology I don´t know how to adjust this. Even when I increase the manufacturing precision there still will be the unsymmetry of the wiring.
Is there a way to adjust the capacity-wise symmetry of the whole setup?
Thanks for any suggestions!
Johannes
