Other Parts Discussed in Thread: FDC2214EVM
While testing the accuracy of the FDC2212 over a range of capacitance, I noticed that the waveform peak (Vpk) seemed to be inversely proportional to the sensor capacitance. For example, at a i_drive current of 146uA and c_sensor of ~60pF, Vpk was measured to be 1.440V. When c_sensor was increased to 260pF, the resulting Vpk was measured to be 1.384V.
In order to compensate for larger capacitance values (~300pf) I attempted to increase Vpk by increasing the setting of i_drive. I noticed that for a given fixed capacitance value, the effect of changing i_drive on the measured frequency DATA_CH0 was significant. The attached plot shows the calculated sensor capacitance (C_tank + C_parasitic) for given drive current settings.
NOTE: The settling count and reference counts are maxed out.
The data shown was taken from a custom pcb of the recommended circuit design, but I was also able to replicate the behavior using the FDC2214EVM eval kit. Basically, as the drive current was increased, Vpk increased but so did the excitation frequency, thus impacting the calculated parallel capacitance. I was able to confirm the change in frequency using an oscilloscope. On the bright side, the frequency measured by the FDC2212 was within 0.05% of the value calculated by the scope.
My main question is: Is this behavior expected? I.e. is the change in measured excitation expected when modifying the drive current? If not is it possible I have made an error durring the device configuration?
My concern is that similar to increasing the drive current, thereby increasing Vpk detected by the sensor, The reduced Vpk caused by increasing sensor capacitance will affect the accuracy of the FDC2212 over the capacitance range. Are there any suggestions that would help negate the effect of decreasing Vpk/excitation frequency values? Also the eval kit gui has an "Detect Idrive Init" button that seems to determine an appropriate i_drive value, is it possible to share how this is done?
Thanks!
-Nathan