Data Noise or Sidebands When Exporting CSV From FlexLogger

Updated Aug 12, 2026

Reported In

Hardware

  • NI-9234

Software

  • FlexLogger

Issue Details

I am acquiring data in FlexLogger using hardware such as the NI‑9234 module.
When I analyze the data directly in TDMS format, the signal looks clean and correct. However, after exporting the data to CSV, I observe unexpected artifacts in the FFT, such as:

  • Sidebands around the main frequency peak with ~240 Hz spacing
  • Apparent noise that is not present in the original measurement
  • Differences between TDMS-based analysis and CSV-based analysis

This occurs when using high-speed acquisition in ~10 kHz range, especially when exporting data manually.

Solution

This issue is caused by a mismatch between the acquisition sampling rate and the CSV export sampling rate. To resolve this issue, follow these steps:

  • Check the acquisition sampling rate in FlexLogger.
  • Find the acquisition rate of the hardware in its specification.
  • Check the CSV export data rate. When exporting to CSV, ensure that the export rate exactly matches the acquisition rate.
  • Set the CSV export rate to the same value as the acquisition rate.
  • Confirm that the sidebands or artificial noise are no longer present.

Use TDMS format as the primary logging format in FlexLogger to avoid these issues entirely.

Additional Information

CSV export in FlexLogger relies on resampling or interpreting the data based on the user-defined export rate. If this rate does not match the actual acquisition rate, the data can be misrepresented in the time domain, leading to. In contrast, TDMS files store the exact timing information and native sampling rate, preventing such inconsistencies.

Using TDMS as the primary file format offers several advantages

  • TDMS maintains the exact acquisition timing, eliminating the risk of rate mismatch.
  • Since no resampling is required, artifacts like sidebands do not occur.
  • TDMS is optimized for streaming large datasets efficiently.
  • TDMS stores additional contextual data that is lost in CSV.