Development of a Compact High-Voltage Functional Electrical Stimulation Device
Haiduo Wang, Ruizhe Tang, Xilin Liu
TL;DR
This work addresses delivering high‑voltage functional electrical stimulation in a compact, wearable form while minimizing patient discomfort. It introduces a switched‑capacitor stimulator coupled with a flyback HV regulator, controlled by an Arduino Nano ESP32, and augmented with a 7‑bit DS4432 IDAC feedback loop to dynamically set $V_{out}$ up to $135$ V and $I_{out}$ up to $20$ mA. A large testing board benchmarks SC performance against conventional H‑bridge stimulators, followed by a compact three‑board wearable prototype achieving a measured rise time of $12.15\ \mathrm{ns}$ (90% of the voltage swing) and an output of up to $135$ V. The design uses off‑the‑shelf components to reduce prototyping cost, and its modular, three‑board approach supports broad rehabilitation applications in wearables and neurostimulation devices, pending safety validation.
Abstract
This report details the design and development of a compact high-voltage functional electrical stimulation (FES) device. Unlike conventional FES systems, the proposed design prioritizes user comfort by leveraging rapid switching times to effectively activate muscles while minimizing stimulation of pain receptors. The device is equipped with a high compliance voltage of up to 135 V, enabling its use across various muscle groups and accommodating users with differing skin conductance levels. At the core of the system is a switched capacitor (SC) stimulator, which facilitates fast switching, and a flyback converter that steps up the battery voltage. The design exclusively utilizes off-the-shelf components, significantly reducing prototyping costs. Initially, a large testing board was developed to evaluate the performance of the SC stimulator in comparison with conventional H-bridge-based current-regulated stimulators. Subsequently, a compact, three-board PCB design was created to enable battery-powered operation, making it suitable for wearable applications. The proposed design methodology is broadly applicable to a wide range of neural rehabilitation applications.
