Solar influence on distance data

[Praneethsvch]

The solar influence on the Maxbotix ultrasonic sensor results in larger distance readings.

A flood captured using distance sensor is a decrease in readings when flooding followed by an increase in the readings due to evaporation. In a post-flood scenario, due to the solar influence ultrasonic might read larger distances indicating flood leaving, which is not true.

Possible solution: Using additional external Temperature compensation to negate the solar effects: HR-Maxtemp

[Praneethsvch]

The following is a graph which compares the solar influence on the ultrasonic readings of two sensors: The one on the left uses internal compensation only and the one on the right uses HR-Maxtemp, an additional temperature compensation which is installed outside the housing in addition to the internal compensation.

Observation:

The standard deviation of the sensor readings is larger when the HR-Maxtemp is used. Also under-compensation has been observed during the day which was not an issue when HR-Maxtemp was not used.

Further we suspect that the solar board or the wind could be contributing to the sensor's noise. Thus, the solar board has been disconnected and and a new test has been running on the battery power only, comparing the effectiveness of HR-Maxtemp.

[Praneethsvch]

Solar board has been disconnected and the two sensors are run on battery for 2 days and here below is the data:

Observations:

• The solar influence has been reduced slightly due to the addition of MaxTemp(right peak is smaller compared to the left), however the sensor data on the right with HR-MaxTemp has more unwanted noise and under-compensation.
• The suspicion of solar panel causing the noise is ruled out

[Praneethsvch]

A new experiment is carried out using a solar shield around the Maxbotix sensor in quest of stable readings and here below is the setup:

Inside view:

The shield has been mounter carefully over the ultrasonic sensor symmetrically and sealed using liquid tape. The following is the sensor data comparing with and without HR-Maxtemp (left is without and right is with MaxTemp):

Observation:

• The noise range due to solar is almost the same compared to when the shield is not used.

[Praneethsvch]

Currently, the input to the Maxbotix sensor is from Feather m0's 3.3V VCC. According to the datasheet, this VCC is regulated by the AK2112 voltage regulator on the board and has an output accuracy of +-1.5%. This is equivalent to a +-49.5mV difference in the Maxbotix's VCC input. Wondering if an additional Cap and a Resistor circuit can be designed to smooth this noise further.

[Praneethsvch]

#22 (comment)