Ruben, We need to get profiles of of your LC tanks' oscillation frequency and amplitude changes over a range of temperatures. Can you put these together for 3-5 of your sensor boards? We have a working theory that temperature is effecting the oscillation characteristics of your sensors, such that the the oscillation threshold is crossed, a false state transition is detected, and therefor the interrupt is fired. This may somehow be aggravated by the calibration routine. We will be conducting some tests to help prove this on Monday morning CST. The idea is to establish temperature transition points at which the oscillation threshold needs to be re-calibrated to compensate for the thermal effects on the sensors. Temperature could be detected by using the internal temperature sensor. Another, less probable, solution is that we find a new oscillation threshold that works across the needed temperature range. Let me know if any of that is unclear.
↧