Open-Source
Multi-Chemistry BMS for Tiny Social Robots (BSc/MSc)
- Survey emerging sodium-ion battery technology and BMS requirements
for small-form-factor devices.
- Design and implement an open-source, multi-chemistry Battery
Management System supporting Na-ion, LiFePO4, and Li-ion 18650
cells.
- Integrate with RSC v5 carrier board and evaluate thermal/safety
characteristics.
- Target OSHWA certification and publish reference design for the
maker/robotics community.
Extended Description
Background
The RSC project requires a reliable, safe, and affordable battery
solution. Sodium-ion technology offers compelling safety benefits (no
thermal runaway, 0V storage capability) and sustainability advantages,
but lacks mature BMS solutions for small devices. This thesis addresses
that gap.
Tentative Research Questions
- What are the technical requirements for a BMS supporting multiple
18650 chemistries (Na-ion, LiFePO4, Li-ion)?
- How might open-source BMS designs (e.g., Libre Solar BMS-C1) be
adapted for compact 3-4S applications?
- What firmware modifications are needed to support Na-ion’s unique
voltage curves and thresholds?
- How does Na-ion perform in a real social robot application (thermal
behavior, runtime, cycle life)?
Deliverables
- Literature Review: Na-ion technology
state-of-the-art, BMS IC landscape, open-source BMS projects
- Hardware Design: KiCad schematic and PCB for 3S1P
multi-chemistry BMS
- BQ76942 AFE integration
- USB-C PD input compatibility
- I2C interface for host MCU telemetry
- JLCPCB PCBA-ready design
- Firmware: (Zephyr RTOS-based firmware) with
configurable (auto-detect?) chemistry profiles
- Evaluation: Bench test characterisation and
integration with RSC prototype
- Documentation: OSHWA-ready documentation, build
guide, and contribution to RSC open-source repository
Skills Developed
- PCB design (KiCad)
- Embedded systems (Zephyr RTOS, I2C bus)
- Battery chemistry fundamentals
- Open-source hardware practices
- Technical documentation
Prerequisites
- Basic electronics knowledge
- Familiarity with KiCad or willingness to learn
- Interest in sustainable technology
Supervision
- Primary: Farnaz Baksh
- Technical: Matevž Zorec (and/or External)
Possible Timeline
| 1 |
4 weeks |
Literature review, BMS IC selection |
| 2 |
6 weeks |
Schematic design, component sourcing |
| 3 |
4 weeks |
PCB layout, PCBA order |
| 4 |
4 weeks |
Firmware development |
| 5 |
4 weeks |
Testing, RSC integration |
| 6 |
4 weeks |
Documentation, thesis writing |
- RSC Gets a HAT (carrier board integration)
- IoT Gateway (power management for always-on operation)
External Collaboration
Potential
- Libre Solar / EnAccess: Upstream contribution to
BMS-C1 project
- OSHWA: Certification process documentation
- Na-ion community: First open-source small-device
BMS reference
Keywords
sodium-ion BMS
open-source hardware KiCad
BQ76942 OSHWA 18650
multi-chemistry Zephyr RTOS
social robotics
Appendix
Notes on RSC
Na-Ion BMS Investigation Summary
Date: January 2026
Context: RSC v5 power subsystem design
Status: Research / Thesis Topic Candidate
Key Question
Should RSC v5 adopt sodium-ion 18650 cells instead of traditional
Li-ion?
Na-Ion Tradeoffs
| Energy Density |
~100-160 Wh/kg |
~250 Wh/kg |
| Safety |
No thermal runaway, can discharge to 0V |
Fire/explosion risk |
| Cost Trend |
Dropping fast (~$59/kWh in 2025) |
~$52/kWh (LFP) |
| Cycle Life |
3,000-10,000+ cycles |
500-2,000 cycles |
| Cold Performance |
Excellent (-30°C @ 92% capacity) |
Poor below 0°C |
| Voltage Range |
1.5V - 4.1V (wider swing) |
2.5V - 4.2V |
RSC v5 Power Constraints
- Target: 4-8 hrs standby/idle
- Form factor: Max 3x 18650 cells (compact desktop
companion)
- Peak draw: 5-9A (CM5 + peripherals under load)
- Input: USB-C PD @ 20V
The Problem
Typical Na-ion 18650 specs (e.g., HAKADI 1500mAh): - Standard
discharge: 0.5C (0.75A) - Max continuous: 3C (4.5A) - Max peak: 5C
(~7.5A)
3S1P = 4.5A continuous → insufficient for 5-9A peak
demand
3S2P = 9A continuous → requires 6 cells (exceeds form
factor!)
BMS Challenge
No dedicated Na-ion BMS ICs exist. Current solutions: 1.
Configurable ICs (TI BQ76952/BQ76942) - programmable
thresholds 2. Smart Chinese BMS boards (JK BMS:
1.2V-4.35V range) 3. Discrete protection circuits - for
simple low-power applications
Open Source BMS Landscape
| Libre Solar BMS-C1 |
BQ76952 + ESP32-C3, KiCad, CERN OHL v2, 16S/100A |
| ENNOID-BMS |
LTC68XX + STM32, modular, EV-scale |
| diyBMSv4 |
Per-cell modules, JLCPCB-friendly |
| Green-BMS |
OSHWA certified (IT000007) |
Gap identified: No open-source Na-ion BMS for small
devices (3S-4S, <10A).
Proposed Contribution
RSC Na-Ion BMS Mini - multi-chemistry open-source
BMS: - 3S-4S support (Na-ion, LiFePO4, Li-ion configurable) - 10A
continuous / 15A peak - BQ76942 AFE + microcontroller - USB-C PD
integration (align with RSC carrier board) - I2C telemetry to CM4/CM5 -
JLCPCB PCBA compatible - OSHWA certification target
Practical Paths for RSC v5
- Accept limitation: 3S1P Na-ion, limit to 4-5A, rely
on USB-C PD for heavy loads
- Larger pack: 3S2P (6 cells) with slightly larger
enclosure
- Wait for 21700 Na-ion: Higher capacity cells
emerging
- Pragmatic hybrid: LiFePO4 now, design BMS for
multi-chemistry future-proofing
References
- Libre Solar BMS-C1: https://github.com/LibreSolar/bms-c1
- HAKADI Na-ion cells: hakadibattery.com
- TI BQ76942 datasheet: ti.com/product/BQ76942
- CATL Naxtra (2025): 175 Wh/kg, 5C charging, 10,000+ cycles