Another variable to consider when designing a system that employs a 3D sensor for angle calculations is the sample rate of the sensor. For example, consider Table 3-1 which applies to TMAG5170.

In the operating conditions presented in Table 3-1, the update rate for the various averaging modes is shown for one, two, or three axis operation.

In any system performing angle calculations on a moving target, the angular step between each measurement will be set by this sample rate. For instance, suppose TMAG5170 is sampling all three axes with the maximum averaging rate, or as we see in Table 3-1, 0.417 ksps. The time between each sample is about 2.4 milliseconds. If the magnet being sensed were rotating at 100 RPM, then the resulting angle change between each sample is 1.44 degrees. This will appear as a fixed offset in the measurement as well as define the minimum detectable change in angle.

Additionally, depending on the sample rate settings, there may be additional angle error. While operating in a pipelined fashion for the maximum sample rate, the typical conversion time for any TMAG5170 conversion is expected to be about 25 microseconds. If averaging 32 conversions, then the total sample time is about 800 microseconds. Over this time frame for the averaging of the first channel, the angle will have changed by 0.48 degrees. By the time the second channel conversion is complete, the angle will have increased by the same amount again for a total motion of about 1 degree. This offset in timing between the two samples will result in angle error. This again highlights the benefit of the pseudo-simultaneous sampling.

Care should be taken in systems where the magnet is rotating at high rates to account for these types of errors and select settings appropriately.