SignalProcessing&ControlBoard

ECE 455 proton magnetometer

One proton precession magnetometer design: a polarization board, a signal-processing and control board with a Raspberry Pi Pico 2 and an ADS1256 ADC, an off-board two-stage amplifier, the sensing and polarizing coils, and the Pico firmware. docs/design.md describes how they fit together and lists open issues.

Layout

Design files, one folder per type:

Everything else:

• util/ reads those files and holds the simulation, breadboard layout, firmware host harness, and tests.
• docs/ has the design description, the verification plan, generated simulation and breadboard reports, datasheets, and references.
• local/ is for uncommitted output.

Setup

python3 -m venv .venv
. .venv/bin/activate
pip install -e '.[test,bench]'
brew install ngspice        # or: sudo apt-get install ngspice

make firmware and make test compile the firmware on the host with $CC, or clang or gcc. Building and flashing the Pico needs the Pico SDK and Arm toolchain; see firmware/README.md.

Commands

make test         # all tests: board/schematic/BOM/firmware consistency, coil forms, LTspice translation, breadboard, estimator
make firmware     # only the firmware estimator tests
make sim          # simulate ltspice/AmplificationSim.asc and regenerate docs/simulation/
make breadboard   # check the breadboard layout and regenerate docs/breadboard/
make verify       # test, sim, and breadboard
make flash        # build the Pico firmware and flash it over USB
make plot         # plot captures from the Pico (sets up .venv on first use)

Status

Both PCBs match their schematics, and the firmware's SPI pins match the ADS1256 header on J3. The amplifier exists only as an LTspice drawing and a breadboard layout; no op amp has been chosen. From the coil workbook and print files, the coils as built should give a best-case FID of about 2 µV, a fifth of the drawing's 10 µV. The coil forms are sized for the sensing coils to sit inside the polarizing coil, where its turn-off would induce on the order of 100 V in them, beyond what the damping switch and amplifier input are rated for. The firmware now polarizes and captures on a button press but does not drive the damping switch. Nothing here has been measured on hardware; see docs/verification-plan.md.