Electrochemical Analyzer.
A tested Python package, GUI, and CLI for analyzing potentiostat voltammetry data — CV, DPV, FS-CV, and Tafel — built to replace the manual macro-plus-Origin workflow I use in lab.
In the Microtechnology Lab I run cyclic voltammetry, DPV, and fast-scan CV on CH Instruments potentiostats, and the analysis workflow is manual: instrument macros, copy into Origin, pick baselines by eye, repeat. ecvolt is my attempt to replace that with something version-controlled and tested.
It’s a Python package with a GUI and CLI covering four techniques. CV: peak detection with proper baseline correction, ΔEp and peak-ratio diagnostics, and Randles–Sevcik fits across a scan-rate series to extract diffusion coefficients. DPV: peak finding, FWHM, two-Gaussian deconvolution for overlapping peaks, and calibration curves with ICH-standard LOD/LOQ. FS-CV: background subtraction, color plots, current-vs-time extraction. Tafel: slope and exchange-current fits in a user-chosen window. The GUI (PySide6) lets you pick baseline anchors and Tafel windows by dragging directly on the curve, but contains zero analysis math — every number comes from a library function, so the same code path is unit-tested headlessly.
The part I care most about is verification. The test suite runs entirely on synthetic data with known answers: the Randles–Sevcik fit recovers a known diffusion coefficient within 5% from deliberately noisy files (exactly from clean ones), Tafel slope and exchange current come back within 2% and 10%, and the demo command reruns byte-identically. Parsing handles the messy reality of instrument exports — unit scaling from column labels, multi-segment scans, deliberately malformed files in the tests.
The next step is validating the parser against my own lab exports and cross-checking results against the existing Origin workflow before trusting it for anything that matters. I have already run some tests against real cyclic voltammetry results, but there is more to do before it can replace the tools I use.