Tool/software: Code Composer Studio
Hello,
in the CapTIvate Design Center it says that if the Conversion Count and Conversion Gain are set to the same value, the minimum amount of offset substraction is applied. In some other document I also read that no offset is applied at all (which does not seem to be true as far as I can tell). Now to my question: Does the "minimum offset amount" depend on the capacitance of the used electrode? I took various discrete capacitors and tried to measure/estimate (in mutual capacitance mode) their capacitances by enabling an internal reference capacitor (or by placing another external cap parallel to the first one) and compare the count to the LTA and any offset would distort the measurement. I assumed following formula for the measurement; with the capacitance of the external cap being the unknown variable:
C(external cap) x LTA = [C(external cap)+C(internal refcap)] x (LTA - Delta)
Up to 8.2 pF the caps could be measured kind of accurately if you keep the conversion count and gain below 250. Lower caps allowed higher count and gain values and some even required them. My larger caps (22 pF and 33 pF) on the other side were impossible to be measured accurately as the calculated values were way too low. I also made sure that there was (hopefully) enough time for the conversion by maxing out the charge and transfer phase lengths. At this point I assumed that there actually might be "large minimal offset subtraction". Maybe the captivate module does not work for these high capacitances and so it subtracts as less as possible (because of equal count and gain values) but as much as needed. Is this assumption true? To somehow verify this I lowered the conversion gain value and there were no changes in the delta and thus the sensitivity, which could mean it applied a large offset before, too, right? For my project the capacitances of my HID are in this range and I am also required to provide the absolute cap values during runtime. I am aware that the MSP430 is not specifically suited for this task, but its amount of channels and the CapTIvate module in general make it very comfortable to use. If I could also acquire information to which "known voltage" the external cap is charged and how large the internal sample capacitor is, I would probably benefit from it.
Also I studied the automatically generated source code and couldn't find any function definitions for the calibration process or for any feature at all. The functions are all declared and I can use them, but I would also like to know how they work. The definitions have to be somewhere, right? Or are they precompiled and I am not even authorized to have access? Sorry if it is a dumb question or if I am too dumb to find what I'm looking for. I am not very experienced, yet. Thank you in advance!
Best regards,
Canbey
EDIT: I found out through debugging that "minimal ui16OffsetTap" is only applied when the ui8CoarseGainRatio and ui8FineGainRatio functions don't suffice to hit those exact conversion count/Gain values that I want. By changing the ui16OffsetTap manually to zero during debug, the pRawCount value changes accordingly. Now if I really want no OffsetTap, I can set that new pRawCount value for Conversion Count and Gain. My initial assumption (see above) therefore is wrong as the GainRatio values were always maxed out because of the high capacitances (22 pF and above), so even lower Conversion Gain values didn't change the sensor's behaviour. You can see this part of my question as solved.
