Tool/software:
As the datasheet and the forum suggests, only channel 0 has a high drive mode. This is something we were not aware when developing the software.
The issue we are having now is, that the LDC doesn't complain when using high drive mode on all four channels. In fact, it even worked without an issue - mostly. For some capacitance-inductance combination we got really nice results when using high drive mode on all four channels. Which is something that shouldn't be possible if I understand your datasheet correctly.
This table shows the IDRIVE values we got when experimentally finding them. Do note that at this point four identical LC sensors were connected - one on each channel. "---" means that we couldn't find a IDRIVE value that worked without the LDC complaining.
Note: C is the capacitance soldered to the coil-PCB. I suspect there is about 150pF of additional parasitic capacitance in the PCB itself. L, R, Q and f are theoretical values calculated by your tool. IDRIVE is defined through experiments.
Another effect that we noticed is on the type of dielectric material used. The table above used cheap X7R capacitors, later on we switched to NP0 (which is better for this application anyway). IDRIVE seems to be lower by around 2 for most value ranges when using NP0 instead of X7R. I'm noting this here, because high current drive worked reliably with X7R, but barely with NP0 - which didn't make too much sense to us. The same coil PCB was used in all experiments.
I guess the question I'm asking is: Are we using undefined behavior here? Is there an error in the datasheet? Why is this working at all?
