DAQ notes

Guides to running MCC, LabJack, Digilent and NI hardware on macOS and Linux, comparisons of DAQ hardware and software, and measurement fundamentals with worked numbers.

Comparisons and buying guides

Choosing the best USB DAQ under $1,000

LabJack, MCC, NI and Digilent sell USB DAQ devices from the $204 MCC USB-1208FS-Plus to the NI USB-6003 (from $928), and they differ most in input ranges and OS support.

6 min read

DAQami alternatives for MCC hardware

TracerDAQ, DASYLab, DAQFactory, MATLAB, LabVIEW and Python can all log Measurement Computing devices, and of these only Python with uldaq runs on macOS or Linux.

7 min read

Data logging software compared

Of the nine DAQ loggers compared here, all but Digilent WaveForms and DAQAtlas (in development) are Windows only, and most are built to sell one vendor’s hardware.

6 min read

LabVIEW alternatives for data acquisition

FlexLogger, DewesoftX, Python with nidaqmx and test platforms like Synnax can take over logging from LabVIEW, but deterministic control on CompactRIO or PXI still runs on LabVIEW Real-Time and FPGA.

7 min read

CSV, MATLAB, TDMS, Parquet or HDF5?

CSV opens in every tool but stores numbers as text, while MATLAB .mat, TDMS, Parquet and HDF5 store binary data and differ in metadata and tool support.

7 min read

DAQ on Mac, Linux and Python

Is there a DAQami for Mac?

DAQami, MCC’s free data logging app, only runs on Windows, but MCC’s open-source uldaq library supports Measurement Computing USB devices on macOS.

6 min read

Using NI DAQ hardware with a Mac

NI-DAQmx has no macOS support and NI-DAQmx Base stopped at macOS 10.14, so NI hardware reaches a Mac through Windows or a PC on the network.

6 min read

Logging data from a LabJack T7 on a Mac

LabJack’s LJM library has native installers for Intel and Apple Silicon Macs, so a T4, T7 or T8 can log to disk from Python or a desktop app without a VM.

7 min read

Data logging with an Analog Discovery 3 on a Mac

WaveForms runs natively on Apple Silicon, and its Record mode streams an Analog Discovery 3 well past the 32,768-sample buffer, at up to about 5 MS/s total to disk.

6 min read

How to open a TDMS file on a Mac

NI’s Excel add-in and DIAdem are Windows only, but a browser viewer or the open-source npTDMS Python library reads .tdms files on macOS and Linux.

6 min read

Running MCC DAQ devices on Linux and a Raspberry Pi

MCC’s open-source uldaq library runs Measurement Computing USB and Ethernet DAQ devices on desktop Linux and Raspberry Pi OS, while the DAQ HATs on a Pi’s GPIO header need MCC’s separate daqhats library.

7 min read

Python data acquisition across DAQ vendors

NI, LabJack, MCC and Digilent all ship official Python packages, but nidaqmx has no macOS support and mcculw runs only on Windows.

6 min read

Measurement fundamentals

How fast should you sample? Sample rate and aliasing

A 900 Hz tone sampled at 1 kS/s lands in your data as a clean 100 Hz sine, and no filtering after the fact can take it back out.

8 min read

Thermocouple measurement with a DAQ

A type K thermocouple produces about 40 µV per °C, so 1 mV of mains pickup on the leads is a 25 °C swing and a CJC sensor 1 °C off moves every reading by 1 °C.

9 min read

Measuring 4–20 mA sensors with a DAQ

A 0.1% 250 Ω shunt turns a 4–20 mA loop into 1–5 V on any voltage-input DAQ, provided the input range still shows the 21 mA fault level.

7 min read

Single-ended vs differential inputs

A grounded source measured in RSE mode closes a ground loop, and a floating source on a differential input drifts to a rail unless it has a bias resistor.

6 min read

ADC resolution vs accuracy

The 24-bit NI-9239 steps in about 1.25 µV, yet its typical accuracy at a 10 V reading is about ±3.8 mV.

7 min read

Hardware-timed vs software-timed sampling

With 100 µs RMS of timing jitter, software-timed sampling limits a 1 kHz signal to about 4 dB SNR, while a hardware pacer keeps sample spacing within its timebase’s ppm rating.

6 min read

How to synchronize multiple DAQ devices

Two DAQ devices rated ±50 ppm can drift 0.36 s apart in an hour, and a shared trigger, a shared clock or a marker signal each removes a different part of that error.

6 min read

Running a mixed-vendor DAQ bench

DAQ devices from different vendors run on their own crystals and write their own file formats, so lining up one run starts with a shared wall-clock start time.

7 min read