PyMEOP

Data acquisition for optical polarimetry of helium-3 polarized by metastability exchange optical pumping (MEOP). PyMEOP scans a probe laser across two absorption lines of the metastable atoms in the discharge, fits the two peaks, and reports the nuclear polarization from the ratio of their heights. It runs as a PyQt5 desktop application on the polarizer's control computer.

Running it

pip install -r requirements.txt
python main.py

On Ubuntu, a Wayland error at startup means qtwayland5 needs installing.

Instruments

Each instrument is reached over the network at the address set in config.yaml:

Setting	Instrument
probe_ip	Toptica DLC probe laser controller (current and temperature)
meter_ip	wavelength meter
lockin_ip	lock-in amplifier reading the probe absorption
siggen_ip	SRS SG380 signal generator driving the discharge

The same file sets the scan timing, the scan x axis (current or wavelength), the baseline under the peaks (baseline_degree: 1 straight, 2 quadratic), and the data and log directories.

Using it

The window has three tabs.

Find Peaks sweeps the probe laser's temperature or current over a wide range to locate the absorption lines and choose the scan range.

Run does the measurement. Each scan steps the laser current over the peaks while reading the lock-in, then fits two peaks on a baseline (app/scanfit.py). The line shape is selectable: Gaussian for Doppler broadened lines at low pressure, or Voigt once pressure broadening is comparable, around 100 mbar. With peak height ratio r and the ratio r0 measured with the gas unpolarized, the polarization is

P = (r/r0 - 1) / (r/r0 + 1)

The zero amplitudes are taken from the latest scan, or averaged over a range of scans selected on the polarization history plot. The Run tab also controls the discharge signal generator, including:

• a Larmor drive, which drives the discharge coil at the helium-3 Larmor frequency for the field entered, optionally swept in frequency, for a set time, then restores the discharge and resumes scanning;
• a discharge-off relaxation cycle, which switches the discharge off and on for set times to measure relaxation.

Slow Controls shows EPICS process variables listed under epics: pvs: in config.yaml, grouped by their group, colored by alarm severity and greyed when disconnected. A PV marked control: true gets a setpoint box and Set button, and must have limits; a value outside them is refused. A readback PV can be shown beside a control's setpoint. If the IOCs aren't reached by broadcast, set ca_addr_list. This uses pyepics, which brings its own Channel Access libraries.

Data

Each scan is written as one JSON line to an event file in data/, holding the raw scan, the fit parameters and uncertainties, the polarization, and under pvs the last value of every slow controls PV when the scan ended (null if disconnected). The file in use is named current_<start>.txt and is renamed to <start>__<end>.txt when closed (times in UTC). Logs go to log/.

Slow controls are also logged on their own, in slowlog/ (slow_dir), named the same way with a new file each UTC day. Each line is a JSON record with a kind:

• snapshot: every PV's last value, IOC timestamp and alarm severity, every log_interval seconds;
• set: a value changed from PyMEOP, with the old and new value and its source (user, scan, larmor, relaxation, range). This covers EPICS controls and PyMEOP's own settings: siggen.freq, siggen.amp, siggen.output, siggen.fm and zero;
• connect / disconnect: a PV came or went.

Data browser

webapp/ holds a separate, read-only web app for browsing and refitting the event files, locally or on a server that mirrors the data (scripts in deploy/). See webapp/README.md.

References

• https://doi.org/10.1016/j.nima.2019.02.019
• https://doi.org/10.1016/j.nima.2021.165590
• https://doi.org/10.1016/j.nima.2023.168792
• https://doi.org/10.1016/j.nima.2025.170870