Feedback Lockin V2

How to use

Each output channel can have feedback either off or on. If feedback is off, you are free to set the amplitude of the output sine curve yourself. This mode mimics a standard lockin amplifier. If feedback is on, you instead set your preferred setpoint. Then, the lockin will try to hold the corresponding input channel at that setpoint by adjusting the output amplitude for you. This can virtually ground contacts, such that you can measure current flowing into them while holding them at fixed potentials.

Requirements

Python 3, numpy, PySide2, pyqtgraph, PyDAQmx

Installation on windows

1. Install python 3 and git.
2. Run pip install pipenv.
3. cd path/to/project.
4. pipenv install numpy pyside2 pydaqmx.
5. pipenv install -e git+https://github.com/pyqtgraph/pyqtgraph#egg=pyqtgraph.
6. Run from here with pipenv run python -m feedbacklockin.

How to run

Run with python -m feedbacklockin (don't forget pipenv if that's how you installed it).

Optionally provide -s path/to/config.ini. For instance, to run locally with no DAQ card, run python -m feedbacklockin -s dev.ini, and to run with the VTI config, use python -m feedbacklockin -s vti.ini.

TCP API

The lockin will start a TCP server listening on the supplied port, or a random port if none is supplied. Connect to it and send commands to control the lockin from external processes. Terminate each command with a newline. Spaces delimit arguments.

• In response to send_data, the lockin will respond with output amplitudes, input voltages, X, phase, and DC offset in an array with Fortran ordering.
• Send set_setpoint CHANNEL VALUE to set the feedback setpoint of the given channel (integer) to the given value (float).
• Send set_amplitude CHANNEL VALUE to set the channel output amplitude if feedback is off.
• Send set_feedback CHANNEL ENABLED to set the feedback setpoint. ENABLED must be 0 for feedback disabled, and 1 for enabled.
• Send autotune to set PID constants.
• Send reset_avg to reset averaging.

Code Overview

The code is located in the feedbacklockin package. We loosely follow PEP8.

main.py creates the main window GUI and ties all the controls to the underlying FeedbackLockin object (defined in fbl.py), which does the heavy lifting. It also initializes the DAQ card and starts up a TCP server if enabled.

fbl.py tracks the state of the lockin. Its most important methods are sine_out, which computes the output voltages for the DAQ (sin_outs.py), and read_in, which computes the next iteration of results given output from the DAQ. The actual amplitude calculations are in lockin_calc.py, averaging is in moving_averager.py, and feedback is in discrete_pi.py and bias_resistor.py.

daq.py and dummy_daq.py should have the same interface. These write out sine curves to the DAQ cards and read in results. dummy_daq.py uses a simulated transfer matrix (tmm.py) rather than talking to real hardware.