Based on the Figure 3-3 hardware circuits, the direct measurement of zero drift and noise in the input of AMC1305M25 and F28377S system can be performed. In this test, the output noise peak-to-peak is evaluated in comparison to different filter types and oversampling ratio multiples. In this test, SDFM output data uses a 32-bit signed-bit binary output and formula (8) for equivalent noise calculations. For complete data, please refer to Figure 5-1, Figure 5-2 and Figure 5-3. Table 4 Zero drift and noise results and equivalent Noise-Free Bit with different filter types and oversampling rate configurations and Figure 3-8 are summarized with the above data.

It can be seen from the test results that higher SINC filter orders or higher oversampling rates yield data with higher resolution and lower noise. Correspondingly, according to formula (7), the system delay increases accordingly. When accurate current and voltage sampling is required, such as 0.1% resolution and precision, at least SINC2 type filter must be selected with an oversampling rate of at least 64. SINC1 or SINC2 filter is a better choice in scenarios that require a fast response and relatively low-precision, such as 1% resolution and precision.

In addition, the combination of different SINC filter orders and oversampling rates yields similar zero drift and noise. Using higher-order filters at this stage results in shorter data latency while maintaining the same resolution and precision. For example, SINC2 OSR128 has a noise level comparable to that of the SINC3 OSR64, but it incurs an additional 64 clock cycles of latency.