← Projects

Instrument control

Plasma Device Control Unit

Open-source control and acquisition unit for plasma experiments.

Plasma Device Control Unit
CAD layout of the aluminum-framed unit ā€” a Raspberry Pi with a stacked HAT, DAC boards, and a front panel of feedthroughs, thermocouple ports, and mains inlet.

An open-source control and data-acquisition unit for my plasma irradiation and hydrogen transport experiments (PIHTI). It runs on a Raspberry Pi with a Python/PyQt interface ā€” logging vacuum and plasma parameters, driving the gas flows, and closing the loop on plasma current ā€” while staying cheap, hackable, and fully documented.

What it does

  • Orchestrates a plasma run end to end ā€” logging chamber pressures and plasma/probe currents through onboard ADC boards over long, steady-state operations.
  • Drives the process-gas mass-flow controllers and closes a feedback loop on the plasma current.
  • Emits sync signals and a front-panel LED to time-align external instruments such as the QMS.
  • Built on a Raspberry Pi with custom PiHat breakout PCBs, Hall current sensors, a proper power supply, and deliberate grounding and EMI proofing.
  • Python software ā€” open-source and documented.

From spaghetti to an instrument

The unit grew up over six years. It began as a bare Raspberry Pi and a HAT living in a plastic food box, wired straight into the experiment. That prototype earned a proper home ā€” an aluminum-extrusion enclosure drawn in CAD, then built, with the boards, breadboard, and feedthroughs laid out behind a real front panel. The current generation adds a proper MeanWell PSU, custom breakout PCBs, and genuine attention to grounding and signal routing: the difference between "works on the bench" and an instrument you can trust.

Where it's going

The next step trades the single box for small, self-contained modules ā€” 3D-printed slide-in cases that drop into a rack, each an ESP32 (or Raspberry Pi) that talks over LAN. Let the network carry the distance, and keep the fast, noisy SPI/IĀ²C buses local to each module, so a sensor's analog front end never has to cross the lab.

It's a deliberate bet against the National Instruments / LabVIEW way: something more flexible, far cheaper, and ā€” honestly ā€” more charming. If there is ever time to finish it. But someone will.

2019 ā€” a bare Raspberry Pi and HAT in a food container, wired straight into the plasma bench.
2020 ā€” the breadboard prototype beside its CAD render and the first aluminum-extrusion enclosure.
2025 ā€” the custom aklab Pi-Hat Breakout (rev 1) PCB, retiring the hand-soldered perfboards for a proper board.
2026 ā€” cleaning up the plug interfaces ā€” the scratched, hand-labeled panel (bottom) gives way to keyed connectors, labeled BNCs, and an RJ45 for LAN (top).
A sibling 3D-printed module ā€” the probe-heater box ā€” small, self-contained, and slide-in, the way the design is heading.
The new direction in the flesh ā€” a networked ESP32 thermocouple logger, 3D-printed, with an OLED readout and phone access over Wi-Fi.