Fire hazard of WHY2025 badge due to 18650 Li-Ion cells

Original Article ↗ Hacker News Discussion ↗

Power requirements vs. badge design choices

  • Many argue two 18650 cells are massive overkill for a conference badge, especially one worn on flammable clothing.
  • Others note the badge runs dual ESP32s, a 4" color LCD with backlight, keyboard, and possibly Wi‑Fi, so current draw and a “weekend-long” runtime likely motivated the design.
  • Some suspect power optimization (sleep modes, efficient backlight use) was traded for simply adding a second cell late in the design.

Parallel 18650 cells and electrical safety

  • There is debate about how hard/dangerous it is to parallel Li‑ion cells:
    • One camp warns that flat discharge curves and mismatched states of charge or health can cause large equalization currents and risk.
    • Another camp says 1S2P is common and safe if voltages are reasonably close at connection; the main danger is the initial parallel-connection event.
  • Concerns are raised about user‑swappable cells in parallel: removing/charging just one and reinserting next to an empty/aged cell can recreate that risky equalization scenario.
  • Some call the unprotected, parallel 18650 design a “fundamentally bad” choice for a mass‑distributed wearable.

Protection circuitry and schematic quality

  • Discussion of the published schematic notes:
    • Parasitic resistances (MOSFETs, traces) and a ~200 Ω “balancing” resistor; some find the design “odd” and hard to interpret.
    • The protection circuit is incomplete (e.g., unclear handling of reverse polarity between cell and protector).
    • Text clutter and confusing annotations (e.g., “LED will burn when battery wrong way round”) are criticized as poor engineering communication.

Cell holders and ‘protected’ cells

  • The chosen holders are likely too short for most protected 18650s, effectively locking users into unprotected cells.
  • This is seen as eliminating the obvious retrofit fix (“just use protected cells”). A few report success with specific shorter protected cells, but compatibility is uncertain.
  • Some prefer custom 3D‑printed holders designed to enforce polarity and fit protected cells.

Alternative chemistries and form factors

  • Suggestions include LiFePO4, NiMH AAs/AAAs, coin cells, or even USB‑only power.
  • Counterpoints:
    • Energy density and max current of CR2032/button cells are likely inadequate for this badge’s display and radios.
    • LiFePO4 reduces thermal runaway risk but still allows very high short‑circuit currents.
    • Flat badges limit AA/AAA options; truly flat rechargeable chemistries (thin NiMH/button formats) are hard to source.

Organizational and process failures

  • Commenters note that an earlier design team advised against these cells and later separated from the event.
  • People attribute the outcome to social/organizational “drama” and management overruling safety‑minded engineers.
  • Several stress that Li‑ion safety should be a first‑order design constraint, not an afterthought.

Badge culture, liability, and expectations

  • Some see this as symptomatic of “one‑upping” badge culture: powerful, flashy hardware rushed to thousands of users with product‑grade risk but hobbyist‑grade process.
  • Others argue lawsuits are unlikely (jurisdiction, advisories issued, no incidents yet), but still feel distributing unsafe gear to the public is unacceptable.
  • Multiple comments generalize: mass‑distributed Li‑ion devices demand professional‑level design, validation, and clear documentation—even for “hacker” events.

General Li‑ion and DIY power‑bank concerns

  • There is broader caution around using unprotected cells, cheap AliExpress boards, and parallel packs in DIY power banks.
  • Recommended safeguards include robust protection ICs, thermal sensing, rigorous worst‑case testing, and avoiding unknown designs where failure analysis and consistency can’t be assured.