RigolWFM

by Scott Prahl

A utility to process Rigol oscilloscope .wfm files

https://colab.research.google.com/assets/colab-badge.svg https://img.shields.io/pypi/v/RigolWFM.svg https://img.shields.io/badge/kaitai-struct-green.svg https://img.shields.io/badge/readthedocs-latest-blue.svg https://img.shields.io/badge/github-code-green.svg https://img.shields.io/badge/BSD-license-yellow.svg

This project is intended to be a comprehensive resource for interpreting waveform .wmf files created by any Rigol oscilloscope. Open source (and Rigol’s own applications) that parse/convert Rigol’s binary .wfm files are sadly balkanized: each program tends to support a single oscilloscope group and the available efforts are spread across a range of languages.

This project leverages a domain specific language (kaitai struct) to represent the binary files. Once a binary file has been described in this text format, parsers can be generated for a wide range of languages (C++/STL, C#, Go, Java, JavaScript, Lua, Perl, PHP, Python, and Ruby).

Documentation can be found at <https://RigolWFM.readthedocs.io>

Using RigolWFM

  1. You can install locally using pip:

    pip install --user RigolWFM
    
  2. or run this code in the cloud using Google Collaboratory by selecting the Jupyter notebook that interests you.

  3. or analyze your files using the kaitai struct IDE (you will need to manually upload the appropriate .ksy file and your .wfm to the IDE). This allows one to interactively reverse engineer binary file formats directly in your browser. This is super helpful for those Rigol .wfm formats that are undocumented or not parsing correctly.

Usage

Once RigolWFM is installed, you can plot the signals from binary Rigol .wfm files by:

import matplotlib.pyplot as plt
import RigolWFM.wfm as rigol

filename = 'example.wfm'
scope = 'DS1000E'

w = rigol.Wfm.from_file(filename, scope)
w.plot()
plt.show()

Alternatively, wfmconvert can be used from the command line. For example, the following should convert all the DS1000E files in the current directory to the .csv format:

prompt> wfmconvert E csv *.wfm

If you just wanted to convert channel 1 from a single file to .csv then:

prompt> wfmconvert --channel 1 E csv DS1102E.wfm

If you wanted to a signal .wav file using the second channel waveform (for use with LTspice) then:

prompt> wfmconvert --channel 2 E wav *.wfm

If you want to create a .wav file with channels one and four as signals (and autoscale for use with Audacity or Sigrok Pulseview):

prompt> wfmconvert --autoscale --channel 14 E wav *.wfm

Status

There is a bit of work remaining (testing, validation, repackaging) but there are binary file descriptions for .wfm files created by the following scopes:

  • DS1000B tested

  • DS1000C tested (two files only)

  • DS1000E tested

  • DS1000Z tested, but with wonky voltage offsets

  • DS2000 tested

  • DS4000 tested

  • DS6000 untested

Resources

This has been a bit of an adventure. In the process of nailing down the basic formats, I have gleaned information from a wide range of projects started by others.

Source code repository

License

BSD 3-clause – see the file LICENSE for details.