Skip to content

Phase 7 — Intensity demo (MMI)

Thump the desk or tap the IMU: after a short warmup, EVENT fires from STA/LTA, the seismogram jumps, and MMI climbs from peak acceleration. No new wires.

As built — TFT showing MMI 10.0 after a hard slam, IMU left, ESP32-S3 right

As built — a hard slam drives MMI to 10, orange bar full, waveform spiking. IMU (left), ESP32-S3 (right), colour-coded jumpers across the star.

Firmware

firmware/07-intensity-ui/ · src/main.cpp

Preconditions

  • Checkpoint 6 passed
  • IMU solid enough that a desk thump couples into it

7.1 Demo

  1. Flash (see 7.2). Screen shows WARMUP while the long-term average fills (~10 s — keep still).
  2. When live: quiet baseline → MMI 0, STA/LTA ~1, flat green seismogram.
  3. Thump the table or tap the IMU.
  4. Ratio spikes → red EVENT; big number is MMI (I–XII). Banner drops quickly.
  5. Hit again anytime — intensity updates with each shake.

Serial: [armed] MMI=… pga=… ratio=… and *** EVENT *** on a trigger.

flowchart LR
  A[Accel ±8g @ 104 Hz] --> B[One-pole HP]
  B --> C["Energy hx²+hy²+hz²"]
  C --> D[STA ~0.5 s]
  C --> E[LTA ~10 s]
  D --> F["ratio = STA/LTA"]
  E --> F
  F --> G[EVENT]
  B --> H[Seismogram Z]
  B --> I[PGA peak]
  I --> J["MMI Wald 1999"]
  G --> K[TFT]
  H --> K
  J --> K
On screen From
Green/red wave High-passed vertical Z
Big number Modified Mercalli Intensity from PGA
EVENT STA/LTA crossed RATIO_TRIGGER
Color bar MMI I … XII

Intensity scale (Wald et al. 1999)

Raw instrumental estimate (no desk attenuation factor):

Wald et al. 1999 (instrumental estimate, no desk fudge factor):

MMI = 3.66 * log10(PGA_cm_per_s2) - 1.66

PGA in cm/s², then clamped to I–XII. Idle below PGA_QUIET_G shows 0.

Sensor PGA ≈ MMI
0.05 g ~IV
0.1 g ~VI
0.5 g ~VIII
1 g ~IX–X
≥ 5.5 g XII

Sensor is ±8 g so XII is reachable (Wald XII needs ~5.5 g). ±2 g would clip near X.

This is a demo, not a seismometer

A breadboard LSM6DSO has no calibration, no seismic mounting, and no ground coupling. The MMI number is a fun illustration of the STA/LTA → PGA → intensity pipeline — it is not a real earthquake measurement. Don't use it for anything that matters.

7.2 Flash

Stop any serial monitor on COM first.

cd firmware/07-intensity-ui
pio run -e esp32-s3-n16r8 -t upload --upload-port /dev/ttyACM0

…
Hash of data verified.
Hard resetting via RTS pin...
========================= [SUCCESS] Took … seconds =========================

pio device monitor -b 115200 --port /dev/ttyACM0

seismo — STA/LTA + MMI (Wald 1999)
MMI = 3.66*log10(PGA_cm/s2)-1.66  XII@>=5.48g  FS=±8g
[armed] MMI=0.0 pga=0.002 ratio=1.2
*** EVENT ***  ratio=18.4  PGA=0.210g  MMI=6.8

7.3 Customize

Edit the CUSTOMIZE block in main.cpp, rebuild/flash.

Want… Change
More sensitive EVENT Lower RATIO_TRIGGER
Fewer false events Raise RATIO_TRIGGER or N_LTA
Idle still shows MMI > 0 Raise PGA_QUIET_G
Banner stays longer Raise EVENT_HOLD_MS
Wave taller Lower WAVE_FULL_SCALE_G

Checkpoint 7

  • [x] Warmup completes; quiet MMI 0
  • [x] Desk thump → EVENT + MMI; hard slam can reach XII
  • [x] Knobs documented

v1 bring-up complete.