Crosstalk

We are using two MAX30101 sensors on the same PCB. Although both devices are configured with identical settings (sample rate, LED timing, integration time, and operating mode), we observe a low-frequency periodic artifact (~4–5 s period) in the measured signals.

The issue appears as a slow sinusoidal-like modulation, which we suspect is caused by a slight timing or clock mismatch between the two sensors, resulting in a beat-frequency-like aliasing effect when the data streams are combined or compared.

We suspect the root cause may be asynchronous sampling, FIFO read timing differences, or internal oscillator drift between the two devices.

Is there any recommended method to properly synchronize two MAX30101 sensors (hardware or software approach), or any known best practice to eliminate inter-device timing drift and avoid this kind of low-frequency artifact?