For any measurement system, it is important to remember the operating signal-to-noise ratio, SNR. In this case, the input referred noise of a Hall-effect sensor, will appear as a spurious magnetic field when observing the output. Ideally, this low level noise will be much smaller in magnitude than the actual input magnetic flux density.

As the input magnitude decreases, the signal-to-noise ratio will suffer, and the result will be less accurate angle measurements.

For many sensors, the only way to counter this effect is to use a stronger magnet or to place the sensor closer to the magnet. Cost and mechanical design may prohibit either of these options. TMAG5170 offers an additional means to reduce this noise further through averaged samples.

User-programmable averaging can be set at values ranging from 2× to 32×. While this does result in longer conversion times, it also allows for a reduction in the total effective noise. Each time the number of samples doubles, the effective RMS value of the input referred noise will scale down by a factor of √2. The benefit of reduced input noise is of particular interest in cases where the input magnetic field is small. This option also can help reduce the impact of other external noise factors such as mechanical vibrations.