LabJack Spectrum Analyzer

Uses the LabJack U6 and python to make a 1- to 4-channel spectrum analyzer.

Prerequisites

python > 3.7, with
- matplotlib
- numpy
- scipy
- tkinter
LabJack Exodriver: https://labjack.com/support/software/installers/exodriver
LabJack python: https://github.com/labjack/LabJackPython

Installing

Download or clone this repository and simply invoke:

python labjacksa.py &

Usage

Overview

The LabJack spectrum analyzer streams data from the LabJack U6 and computes its power spectrum using scipy, applying a Hann window to eliminate edge transients. The U6 can stream data at up to 50 kHz, multiplexing the ADC between channels: a single channel can be sampled at up to 50 kHz, two channels at 25 kHz, and so on. Sampling at higher frequencies means a higher frequency cutoff for the power spectrum; sampling for longer times moves the lower limit of the power spectrum to lower frequencies.

The user interface presents the following main areas (from top to bottom):

a menu bar for file review and data-streaming settings
the matplotlib navigation toolbar
a matplotlib plot with showing a time series and its power spectrum
a toolbar to select channels, and control acquisition and data saving
a status bar

Menu bar

Open - opens a data file for review.
Freq - sets the per-channel sampling frequency. Must be set <= 50 kHz / number of channels.
Time - sets the sampling time. This represents the minimum sampling time. The U6 streams data in packets, and the requested sampling time may represent an non-integer number of packets; the actual sampling time may be longer, as we round up the number of packets to the next highest integer and do not discard any of the last packet.
Scaling - sets the units and scaling prefactor. MathTeX may be used for formatting the units string. For example, if sampling an accelerometer + amplifier with a sensitivity of 0.1 m^2/s per volt, set the unit to "m$^2$/s", and the prefactor to 0.1.
About - displays a copyright and license notice.

Acquisition toolbar

Channels are selected using the checkboxes on the left side of this bar. Remember to limit the sampling rate appropriate to the number of channels selected: if it is too high, the status bar will display a message to remind you.

Start and Stop buttons start and stop data acquisition.

Save last saves the currently displayed data to a file.

Setting the save all check box will display a folder-select dialog. Use this to choose an existing folder, or enter a name to create a new folder, and click OK. Data will then be saved in timestamed files until the save all check box is cleared.

Saved data format

Saving data dumps the raw (unprocessed, unscaled) channel data to a file as a JSON object, along with sampling information, and the scaling and units used for display. The JSON object has the following key/value pairs:

prefactor - the data scaling prefactor
scaling - the data scaling unit
rate - the effective sampling rate (i.e. 1/time between points in each channel)
points - the number of points in each channel
dropped - the number of points dropped due to buffer over-runs
channels - a mapping of each active channel name to its data as a list of floating point numbers.

The accelerometer example given above would produce a file that looks like the following:

{"prefactor": 0.1, "unit": "m$^2$/s", "rate": 50000, "points": 100800, "dropped": 0, "channels": {"AIN0": [0.0017877728041639784, 0.0014722472351422766, 0.0021032983731856802, 0.002418823942207382, 0.0017877728041639784, 0.0014722472351422766, 
... ... ...
-0.005468447568546253, -0.005152972058112937, -0.004522021037246304, -0.0048374965476796206]}}

Authors

Mick Phillips - MickP

License

This project is licensed under the GPL v3 License - see the LICENSE.md file for details

Acknowledgments

Thanks to LabJack for making a high-quality, low-cost data acquisition system with a well-supported python interface.

mickp / ljsa