Bandit Dev Platform

Bode Bandits Analysis Tool

The Bode Bandits Analysis Tool provides frequency response characteristics of a multitude of filters. A Python GUI allows interface with the tool, providing configuration and visual representations of the measured transfer function.

Hardware

Board and assembly files are provided in the repository, optimized for JLC's PCBA service.

Methods

The Least Mean Squared (LMS) adaptive filter methodology is implemented in a semi-real time fashion, allowing rapid characterization cycles to be made.
All algorithms (LMS, FFT, Downsampling) are fully parametric, and largely user selectable on the fly. Automatic self-calibration corrects for
component tolerance variations, VDDA magnitude, and the frontend anti-aliasing filter transfer function. DC calibration routines are run at
startup, all transfer function calibration metrics are gathered at startup as well as upon change of any parametric setting.

Real Time Adjustment Parameters

• Range of frequency analysis (set ranges)
• Datapoint resolution (2^n)
• Acceptable LMS convergence error tolerances

Required User Software

• Python 3.9+
• That's it :)

Repo Layout

• App: Contains the PC side of the project. Including the Python source and helper tools.
• HDL: Placeholder for potential HDL in the event a switch to a more PL based approach occurs.
• src: All C source for both the current main build as well as helper C files and feature tests.
• Documentation: All standard documentation as well as theory behind every working component, with test programs and scrips included.