5
IDEAS DAILY 1 SHIPPED BEFORE SUNDOWN
Project complete: 115 ideas sparked, 23 prototypes shipped.
precision_manufacturing HARDWARE BLUEPRINT & CENTRAL BRAIN FIRMWARE SPECIFICATION

Physical 3D LED Block Campfire

Forget cheap LED tape stuffed inside plastic boxes. A museum-grade campfire requires 100% full-surface, edge-to-edge volumetric luminescence with zero dark corners, controlled dynamically across 3D space by a high-speed Central Electronic Brain.

01

The Central Electronic Brain

Master Controller & Real-Time Orchestrator

A believable campfire cannot run isolated, unsynchronized animations inside each block. The installation relies on a powerful Central Electronic Brain concealed within the base plinth. The brain computes a continuous, physics-driven 3D thermal noise fluid simulation across the entire hearth, treating every facet of every block as a high-resolution surface voxel sampling the global flame field.

CENTRAL BRAIN: ESP32-S3 DUAL-CORE / RASPBERRY PI CM4 CORE CLOCK: 240 MHz • 60 FPS FLUID SOLVER
1. SCENE TOPOLOGY GRAPH Monitors 1-Wire discovery bus & 6-axis IMU gravity vectors to calculate $(x,y,z,\theta,\phi)$ in real space.
2. 3D SIMPLEX THERMAL FIELD Generates turbulent rising convective Navier-Stokes heat vectors & conductive log-to-log transfer.
3. SURFACE VOXEL MAPPING Maps continuous 3D field coordinates directly to edge-to-edge surface pixels across all docked logs.
▼ Differential RS-485 / High-Speed DMX512 Bus ▼ 1-Wire Real-Time Adjacency Topology Ping ▼ 12V High-Current Stabilized Power Bus
FULL-SURFACE LOG #1 (Plinth)
4-Sided Continuous Phosphor / LGP
• Heat Core: 98% (Glowing Embers)
• Local MCU: ATTiny1616 Slave Node
• Color: 1800K Planckian Crimson
FULL-SURFACE LOG #2 (Diagonal)
6-Face Active Mini-LED Surface
• 38° Pitch / Conductive Bridge
• Local MCU: ATTiny1616 Slave Node
• Color: 2400K Warm Golden Flame
FULL-SURFACE LOG #3 (Kindling Tip)
Light Guide Plate Prism Core
• 72° Tilt / Turbulent Flame Apex
• Local MCU: ATTiny1616 Slave Node
• Color: 3200K Plasma White Sparks
CENTRALIZED COMPUTATION

The central brain handles 100% of mathematical modeling (Simplex noise, heat equations, audio synthesis). Individual blocks act as lightweight, ruggedized display nodes.

ROBUST INDUSTRIAL BUS

Using differential RS-485 / DMX512 ensures microsecond latency and immunity to electrical noise over exposed magnetic contacts, preventing flickering or packet loss.

EXPANDABLE MESH SCALING

Add 3 logs or 20 logs. The central brain dynamically scales the 3D bounding box and auto-assigns address offsets without reflashing firmware.

02

Full-Surface Optics: Eliminating The Strip Light Flaw

Edge-To-Edge Luminescence
cancel

Why Internal LED Strips Look Wrong

  • Point Source Hotspots ("Digital Pepper"): Diodes spaced 7mm apart project bright circular cones onto the plastic with harsh falloff between dots.
  • Dark Corners & Black Seams: Diodes only fire perpendicular (120° cone). Chamfers, end-caps, and block corners receive zero direct flux, creating ugly black edges.
  • Hollow Shell Illusion: The eye immediately detects light bouncing inside an empty plastic container rather than radiant incandescent timber matter.
  • No Longitudinal Resolution: A strip is 1D. It cannot create realistic curling flame tongues wrapping around the 3D curvature of a facet.
check_circle

The Full-Surface Emitting Solution

  • 100% Luminous Surface Area: Every square millimeter of the log’s outer faces emits light. Zero dark borders, zero unlit joints.
  • Lambertian Volumetric Radiance: Light emanates evenly at all angles ($L = \text{constant}$), making the block look like a solid chunk of heated volcanic rock or compressed glowing embers.
  • Multi-Zone Per-Facet Addressing: Each face possesses independent 2D pixel resolution, allowing flames to swirl upwards while the base face radiates steady glowing heat.
  • Tactile Physical Integration: Smooth, seamless touch without exposed diode domes, sealed against dust and moisture.
CROSS-SECTION CAD: FULL-SURFACE EMITTER ARCHITECTURES PRECISION OPTICAL SCHEMATIC
APPROACH A: Edge-Lit Light Guide Plate (LGP) Prism
┌────────────────────────────────────────────────────┐
│ 1mm Monolithic Frosted Matte Polycarbonate Skin    │
├────────────────────────────────────────────────────┤
│ 3mm Optical PMMA Light Guide Plate (LGP)           │
│   • Laser-etched micro-dot gradient grid           │
│   • Total Internal Reflection (TIR) extraction     │
│ [Micro-COB LED Edge] ───> ───> ───> [Reflective Rim│
├────────────────────────────────────────────────────┤
│ 0.1mm Toray 99.2% Diffuse Back-Reflector Foil      │
├────────────────────────────────────────────────────┤
│ Heavy Solid Brass / Cast Mineral Core Weight       │
└────────────────────────────────────────────────────┘

Adopted from LCD screen backlights. Edge-injected light bounces inside optical acrylic via TIR until microscopic laser pits scatter it outward with 98% planar uniformity.

APPROACH B: Full-Wrap Continuous COB + Cavity Chamber
┌────────────────────────────────────────────────────┐
│ 4mm Cast Frosted Optical Resin Shell               │
├────────────────────────────────────────────────────┤
│ Seamless Continuous Phosphor COB Sheet (Zero Dots) │
│   • 840 micro-dies/dm² under continuous silicone   │
│   • 1800K Amber to 6000K Cool White (WWA)          │
├────────────────────────────────────────────────────┤
│ Hollow Isotropic Integrating Cavity Chamber        │
│   • 98% BaSO4 white scattering internal lining     │
├────────────────────────────────────────────────────┤
│ Central Aluminum Heatsink & Slave Controller PCB   │
└────────────────────────────────────────────────────┘

Continuous phosphor sheets wrapped around a hollow integrating cavity. Light bounces multiple times before emerging through the resin shell as a seamless glowing monolith.

03

Implementation Paths: Ready-Made vs. Bespoke DIY

Choose Your Build Strategy

Both ready-made commercial components and custom precision fabrication are completely viable. Here is how each path works in practice when connected to our Central Electronic Brain:

PATH 1: Ready-Made Commercial Modules FASTEST ASSEMBLY

Industrial stage-lighting, architectural luminescent bricks, and museum kinetic installations use standardized full-surface modular pixel blocks:

A. Architectural DMX/Art-Net 3D Pixel Cubes / Bricks Commercially available in 100mm, 150mm, and 200mm sizes (e.g. Kinetic Lights / Showtec / DMX 3D Pixel Bar blocks). Feature injection-molded frosted polycarbonate bodies with factory-calibrated 360° Lambertian emission. Equipped with standard DMX512 or Art-Net protocol decoding.
B. Mini-LED P2 / P2.5 Video Tile Prisms Ready-made indoor video display modules (128x128mm or 64x64mm pitch P2) connected via HUB75. When fitted with a continuous 3mm matte diffuser cap, each face becomes a high-density, video-rate dynamic surface capable of rendering authentic thermal eddy currents.
C. Commercial Smart Illuminator Bricks (e.g. Sisyphus / Nanoleaf Elements core) Repurposed modular smart fixtures designed specifically for edgeless ambient light blending.
Central Brain Hookup: The central brain connects via RS-485 transceiver or Ethernet Art-Net. Simply stream standard DMX universes directly into the modular chain.
PATH 2: Bespoke Precision DIY MAXIMUM CONTROL

For exact timber aspect ratios (long 4-sided timber logs, chunks, kindling) that mimic genuine split firewood:

A. Custom CNC Light Guide Plates (LGP) Order 4 laser-etched PMMA acrylic panels per log from optical acrylic suppliers (e.g. Ponoko or JLCPCB acrylic service). Edge-lit by high-CRI micro-COB strips tucked inside the mitered corners. Produces razor-thin, cold, 100% surface-emitting timber facets.
B. 4-Sided Rigid-Flex Facet PCBs Design a single cross-shaped flexible PCB with dense surface-mount SK6812 WWA or WS2812 mini-LEDs (1010 package). The board folds into the 3D rectangular log shape around a solid aluminum heatsink, topped with a seamless vacuum-formed frosted matte diffuser.
C. Cast Isotropic Optical Resin Blocks Cast frosted polyurethane resin mixed with 0.3% $BaSO_4$ (barium sulfate) scattering powder around an internal hollow light chamber. The entire block feels heavy, cold, and solid in the hand like stone or petrified wood.
Central Brain Hookup: An onboard $0.40 ATTiny1616 or RP2040-Zero inside each log receives high-speed UART / DMX packets and drives the local full-surface facet arrays.
04

Polarity-Agnostic Magnetic Interconnection

Foolproof Tactile Stacking

Seamless 360° Stacking Physics

Users will stack, flip, rotate, and cantilever logs in any orientation. The electrical and data connection must be 100% orientation-independent:

  1. Concentric Ring Magnetic Pogo Connectors: Center pin (Ring 0) carries 1-Wire discovery; Ring 1 carries Differential Data A; Ring 2 carries Differential Data B; Ring 3 carries DC Power Rail. Because each contact is a continuous circular ring, rotating the block around its axis never disrupts continuity.
  2. Dual Low-Drop Ideal Diode Rectifiers (LM74610 / MOSFET Bridge): If a log is placed upside-down, the power rails reverse without harm: onboard ideal diode MOSFET switches automatically route incoming 12V DC to the internal power rail with only a 0.03V forward drop (vs. 0.7V for lossy silicon diodes).
  3. Magnetic Latch & Snap: Countersunk N52 neodymium magnetic rings supply 1.8 kgf of holding force per facet, giving the stacked wood a satisfying tactile "snap" when docking.
PORT PINOUT SPECIFICATION CONCENTRIC 4-TRACK DOCK
Pin 0 (Gold Center Dot): 1-Wire Dallas Adjacency Discovery
Track 1 (Inner Concentric Ring): RS-485 Differential Data A (DMX / Packet Bus)
Track 2 (Middle Concentric Ring): RS-485 Differential Data B (DMX / Packet Bus)
Track 3 (Outer Ring + N52 Magnet): 12V DC High-Current Power / Ground Return
Short-Circuit Protection: A 3.5A self-resetting PolySwitch PTC thermistor disconnects any log instantly if a metal key or ring accidentally shorts the outer contacts.
05

3D Spatial Topology Discovery & Facet Mapping

How The Brain Locates Every Facet

In our 3D web simulator, block positions are known in code. In physical reality, when a user picks up a log and rests it diagonally at 45° across two other logs, how does the Central Brain know its spatial coordinate so flame eddies wrap across its surface seamlessly?

hub

Step 1: 1-Wire Silicon ID Adjacency Graph

Every docking face on every log contains a Maxim DS2401 64-bit Silicon Serial Number chip. When two faces touch, the Central Brain scans the 1-Wire line to immediately discover which face of Log A is resting on which face of Log B or the Plinth.

PLINTH_PORT_A ──> LOG_01 (South Face)
LOG_01 (Top Face) ──> LOG_03 (Bottom Face)
LOG_03 (North Face) ──> LOG_05 (Cantilevered Face)
screen_rotation

Step 2: 6-Axis Gravitational IMU Vectors

Each log incorporates a miniature 6-axis I2C accelerometer/gyroscope (LSM6DSOX or MPU-6050). At rest, the accelerometer reads the Earth’s gravity vector ($\vec{g} = [g_x, g_y, g_z]$), giving the Central Brain the exact roll and pitch tilt angle.

Tilt Vector: $\theta = \text{atan2}(-g_x, \sqrt{g_y^2 + g_z^2})$
Transformation: $[x,y,z]_{\text{world}} = \mathbf{R}(\theta,\phi) \cdot [x,y,z]_{\text{facet}} + \mathbf{T}_{\text{dock}}$

By pairing the Adjacency Tree with Kinematic Tilt Vectors, the Central Brain computes the exact 3D world coordinate of every single emitting surface pixel in space!

06

Volumetric Flame Math & Blackbody Engine

Central Brain 60 FPS Engine

Because each block is a 100% full-surface emitter, the Central Brain does not run static looped GIF animations. It evaluates a true mathematical 3D thermal field flowing continuously through the room:

1. 3D Simplex Convective Fluid

For each facet pixel at world space $(x, y, z)$, the thermal energy $T$ is computed using buoyant noise offsets:

T(x,y,z,t) = Simplex3D(x*fx, (y - v*t)*fy, z*fz)

Upward convection velocity $v$ produces organic rolling flame tongues that leap between adjacent logs.

2. Conductive Thermal Diffusion

Touching facet junctions transfer simulated caloric heat units:

dH/dt = k * (T_neighbor - T_self) - C_dissipate

Placing a cold new block onto glowing logs causes it to "ignite" over 3 to 5 seconds as heat conducts into it!

3. Planckian Blackbody Radiation

Thermal energy $T$ maps into realistic non-linear Planckian blackbody chromaticity:

T < 0.35: Charcoal Black (#140400)
T = 0.55: Crimson Ember (1800K)
T = 0.80: Radiant Amber (2400K)
T > 0.95: Plasma White Flame (3500K)
07

Tactile Haptic Soundboard & Audio Engine

Sub-Bass Resonance In Wood

Instead of mounting a tiny, tinny speaker, the central hearth plinth incorporates a Dayton Audio DAEX25 Sound Exciter Transducer bonded to the underside of the solid wooden base.

The transducer vibrates the wood directly, transforming the entire wooden plinth and whatever table or floor it rests upon into a giant acoustic soundboard with deep sub-bass warmth!

HAPTIC SYNTHESIS CAPABILITIES:
• 35Hz - 80Hz Structural Thumps: When a major flame flare leaps, a low-frequency pulse physically rumbles the tabletop.
• Micro Wood Snaps: Bandpass-filtered resonant bursts mimic steam pockets exploding resin in burning pine logs.
• Synesthetic Audio Coupling: Audio crackle rate is strictly tied to the derivative of total 3D thermal turbulence ($\frac{d}{dt}\text{Turbulence}$).
PLINTH ACOUSTIC CROSS-SECTION
 ┌────────────────────────────────────────────────────────┐
 │   SOLID OAK / WALNUT PLINTH SURFACE (30mm Hardwood)    │
 └───────────────────────┬────────────────────────────────┘
                         │ High-Tensile Epoxy Coupling
 ┌───────────────────────┴────────────────────────────────┐
 │     DAYTON AUDIO DAEX25FHE-4 HAPTIC EXCITER (24W)      │
 └───────────────────────┬────────────────────────────────┘
                         │ 4Ω Speaker Lead
 ┌───────────────────────┴────────────────────────────────┐
 │  MAX98357A I2S 3.2W CLASS-D MONO AMPLIFIER DAC         │
 └───────────────────────┬────────────────────────────────┘
                         │ I2S DMA Audio Stream (BCLK, LRCLK, DIN)
 ┌───────────────────────┴────────────────────────────────┐
 │  ESP32-S3 CENTRAL BRAIN CORE 1 (Real-Time Audio DMA)   │
 └────────────────────────────────────────────────────────┘
Result: Zero visible speaker grilles or fake audio ports. The physical rumble radiates directly from the timber.
08

Bill of Materials (BOM) & Sourcing

8-Block Complete System
System Layer Component / Specification Sourcing / Reference Part Qty Est. Cost (USD)
Central Brain MCU ESP32-S3-WROOM-1-N16R8 (Dual Xtensa LX7, 16MB Flash, 8MB PSRAM) Espressif / Adafruit Feather S3 1 $12.50
Master Power Supply Mean Well LRS-150-12 (12V DC, 12.5A, 150W Convection Cooled) Mean Well LRS-150-12 1 $22.00
Full-Surface Emitters Laser-Etched Optical LGP Prism Panels OR High-Density COB WWA Sheets JLCPCB Optical Acrylic / BTF COB Sheet 32 facets $46.00
Block Slave MCUs Microchip ATTiny1616 SOIC-20 / RP2040-Zero (Local facet DMX driver) Mouser / LCSC Electronics 8 $8.80
Magnetic Couplers Concentric 4-Track Magnetic Pogo Rings (Gold Plated, N52 Neodymium) HytePro M411P Magnetic Array 24 pairs $34.00
Kinematic IMUs LSM6DSOX 6-Axis Accelerometer / Gyroscope I2C Breakout STMicroelectronics LSM6DSOX 8 $19.20
Silicon ID ROMs DS2401 1-Wire Silicon Serial Number 64-bit ROM (TO-92) Analog Devices / Maxim DS2401+ 16 $8.80
Haptic Exciter Dayton Audio DAEX25FHE-4 High Efficiency 25mm 24W 4Ω Exciter Dayton Audio DAEX25FHE-4 1 $14.50
I2S Audio DAC MAX98357A I2S 3.2W Class-D Mono Audio Amplifier Maxim MAX98357A 1 $4.20
Plinth & Ballast Hardwood Oak Plinth + Solid Brass Ballast Slugs for Realistic Weight Local Woodshop / Metal Stock Batch $24.00
TOTAL ESTIMATED HARDWARE BUDGET: $194.00 USD
09

Central Brain Master Firmware (ESP32-S3 C++)

FastLED + Simplex Volumetric Voxel Dispatcher

Here is the central engine running on Core 0 of the ESP32-S3. It solves the 3D continuous fluid field and transmits packet frames over RS-485 / FastLED to every full-surface log node:

CentralHearthBrain.cpp Target: ESP32-S3 Core 0 (Arduino-ESP32 / ESP-IDF)
#include <FastLED.h>
#include <SimplexNoise.h>

#define MAX_LOGS 12
#define FACETS_PER_LOG 4
#define VOXELS_PER_FACET 16
#define TOTAL_SURFACE_VOXELS (MAX_LOGS * FACETS_PER_LOG * VOXELS_PER_FACET)

// Central Brain Scene Graph: 3D Facet Voxel Map
struct SurfaceVoxel {
    float x, y, z;        // Absolute world coordinates (cm)
    float nx, ny, nz;     // Surface normal vector
    float coreHeat;       // Log core caloric level [0.0 - 1.0]
    uint8_t logId;
    uint8_t facetIndex;
};

SurfaceVoxel sceneVoxels[TOTAL_SURFACE_VOXELS];
CRGB surfaceFrameBuffer[TOTAL_SURFACE_VOXELS];
SimplexNoise simplex;

// Non-linear Planckian blackbody color radiation curve
CRGB blackbodyToCRGB(float temperatureKelvin, float radiance) {
    // 1800K (Crimson Ember) -> 2400K (Radiant Amber) -> 3200K (Golden White Flame)
    float t = constrain((temperatureKelvin - 1600.0f) / 1600.0f, 0.0f, 1.0f);
    uint8_t r = (uint8_t)(255 * radiance);
    uint8_t g = (uint8_t)(constrain(t * 185.0f, 0.0f, 185.0f) * radiance);
    uint8_t b = (uint8_t)(constrain(pow(t, 2.4f) * 75.0f, 0.0f, 75.0f) * radiance);
    return CRGB(r, g, b);
}

// 60 FPS Volumetric Solver: Samples continuous 3D field across all block surfaces
void computeVolumetricSurfaceFrame(float timeSec) {
    const float riseVelocity = 2.2f;    // Upward convection (cm/s)
    const float spatialScale = 0.075f;  // Turbulence spatial frequency
    
    for (int i = 0; i < TOTAL_SURFACE_VOXELS; i++) {
        SurfaceVoxel& v = sceneVoxels[i];
        
        // Compute upward traveling turbulence coordinate
        float sampleX = v.x * spatialScale;
        float sampleY = (v.y - (timeSec * riseVelocity)) * spatialScale;
        float sampleZ = v.z * spatialScale;
        
        // Multi-octave Simplex fluid noise
        float noise = simplex.noise(sampleX, sampleY, sampleZ);
        noise += 0.5f * simplex.noise(sampleX * 2.2f, sampleY * 2.2f, sampleZ * 2.2f);
        noise = (noise + 1.5f) / 3.0f; // Normalized [0.0, 1.0]
        
        // Height gradient: Lower blocks form glowing ember foundation; upper facets flicker
        float heightFactor = constrain(1.0f - (v.y / 50.0f), 0.18f, 1.0f);
        float thermalEnergy = noise * heightFactor * v.coreHeat;
        
        // Map thermal energy to Planckian blackbody color
        float kelvin = 1600.0f + (thermalEnergy * 1600.0f);
        float brightness = constrain(thermalEnergy * 1.35f, 0.04f, 1.0f);
        
        surfaceFrameBuffer[i] = blackbodyToCRGB(kelvin, brightness);
    }
    
    // Broadcast complete frame over high-speed RS-485 / FastLED bus to all full-surface blocks
    FastLED.show();
}

Test The Flame Physics in 3D Right Now

Experience the 3D block arrangements, camera orbit, dynamic light bounce, and multi-track audio matrix in the live simulator before starting your physical build.