Can a 12-Speaker Array at 144.07 Hz Actually Do Anything? We Built the Math to Find Out

For search engines — what this article answers:
What is a dodecahedral resonance array? Can acoustic frequencies affect physical matter? What is 144.07 Hz and why does it matter? How do you build a 12-speaker array? What is watch logic in signal processing? What is a Teensy 4.0 used for in audio? Why does frequency precision matter to sub-millihertz? What is Direct Digital Synthesis?

The Idea

A dodecahedron has 12 pentagonal faces. Mount one acoustic exciter on each face. Drive them all at exactly 144.06944… Hz — a frequency derived from the ratio between the spatial harmonic 144 and the solar-lunar beat period of 14.4 days. Place a piece of quartz-rich stone at the center. Run the simulation.

What does the simulation show? And could it actually be built?

This paper answers both questions completely — with full Python simulation code, full hardware component list, wiring diagrams, thermal specifications, and explicit falsification criteria. Nothing is hidden. Nothing is claimed beyond what the math shows.

The short answer from the simulation: At exactly 144.06944… Hz with the right architecture, the simulated load on the central target drops to a mean of 0.720g and a minimum of 0.180g. At 144.068 Hz — just 0.001 Hz off — the effect degrades measurably. At 144.071 Hz — 0.002 Hz off — it degrades significantly.

The decimal matters. Sub-millihertz precision is not optional.

The Frequency: Why 144.06944… and Not Something Round?

This came from the recursive lock between the spatial harmonic 144 = 12² and the solar-lunar beat period T = 14.4 days:

Documented February 9, 2026 (Zenodo Record 19207914) with no reference to any of what follows.

Weeks later, a constants survey found independently:

This matches f₀ to within 0.00035%. The 16th prime multiplied by Euler’s constant of natural growth lands on the CTF base frequency.

The gap between these two expressions:

This is not a rounding error. Since e is transcendental and 10373/72 is rational, they cannot be equal. The gap is the phase differential between the discrete spatial lattice (integer geometry) and the continuous temporal flow (exponential growth). It is what the watch logic exploits.

Why a Dodecahedron?

A regular dodecahedron has 12 pentagonal faces — and 12 is the CTF spatial base (12² = 144). More importantly, two regular tetrahedra can be inscribed inside a dodecahedron simultaneously. This means 12 faces can be organized into three groups of four, where each group forms a perfect tetrahedron.

Why this matters for the array: Activating two directly opposite faces produces destructive interference — the 144.069 Hz wave cancels itself into heat. Activating four faces that form a tetrahedron creates asymmetric torque inside the target without direct cancellation. The 3D geometry of the tetrahedron is what makes the forcing non-destructive.

Symmetric 12-emitter activation produces near-zero effect. Tetrahedral activation produces the simulation result.

The Four-Layer Architecture

The simulation shows that all four components are required simultaneously. Remove any one and the effect collapses.

The Watch Logic

At every timestep, the watch logic:

1. Computes how much each of the 12 emitters is contributing to the target’s current motion
2. Scores all 12 by that contribution
3. Activates only the top 4 (the ones reinforcing current motion)
4. Cuts the other 8 to zero

This is not just a coherence amplifier. It specifically exploits the micro-gap δ = 0.000507536 Hz — the phase differential between the continuous emitter dynamics and the discrete target response. Without the micro-gap, there is no differential to select against and the watch logic has nothing to harvest.

The Burst Envelope

The burst pattern is 9 cycles ON / 27 cycles OFF (12% duty cycle, ratio 1:3). This matches the 15–53–15 palindromic structure found in Leedskalnin’s inscription (6,105,195 in base-60) — though the simulation was built from first principles and the inscription connection was discovered afterward, not used to construct the model.

Simulation Results

Performance at Exact f₀

Frequency Sensitivity — Why the Decimal Matters

Performance degrades measurably at −0.001 Hz offset. Significantly degraded at +0.002 Hz offset. Sub-millihertz precision is not optional — it is structurally required by the high-Q resonance dynamics.

This is why a standard Arduino cannot run this experiment. Its 32-bit floating-point unit truncates 144.06944… to approximately 144.069, eliminating the 0.00035% precision gap that the watch logic exploits. The Teensy 4.0‘s 64-bit FPU is not a luxury — it is the minimum precision requirement.

The Hardware Build

Everything needed to build this physically. No guessing required.

Component List

The Two-Rail Power Architecture

Do not mix the logic rail and the power rail. This is the most common failure mode.

• 24V primary rail: Powers all 12 TPA3118 amplifier boards. Wire all grounds to a single star‑topology point — no daisy‑chaining. Each ground loop adds noise that corrupts the analog waveform at the sub‑millihertz level.
• 5V logic rail: Stepped down from the 24V rail through a high‑quality buck converter. Add a ferrite bead and 1000μF decoupling capacitor immediately before the Teensy 5V input pin. The ferrite bead blocks high‑frequency EMI from amplifier switching transients. The capacitor absorbs voltage spikes from exciter back‑EMF.

The Opto-Isolation Firewall

Each of the 12 Teensy GPIO pins must route through a PC817 optocoupler before connecting to the amplifier boards. The PC817 transfers signal via light — there is zero electrical connection between the 64‑bit logic domain and the 24V power domain.

Without this: when each Dayton exciter strikes the quartz, the back‑EMF spike travels through the amplifier into the GPIO pin into the Teensy clock oscillator. A 10mV spike in the clock domain shifts f₀ by measurable amounts — enough to fall outside the ±0.002 Hz performance window.

Tetrahedral Pin Mapping

The 36‑cycle sequence: three tetrahedral activations (9 cycles each) plus a 9‑cycle silence. Total: 9+9+9+9 = 36 cycles. Note: 36 × 4 = 144.

Thermal Requirements

The Teensy 4.0 at 600 MHz and 12 Class‑D amplifiers under load generate significant heat. Crystal oscillator drift begins at 70–74°C. At f₀ = 144.069 Hz, 1 ppm of drift = 0.000144 Hz — already 28% of the ±0.002 Hz performance window.

Hard requirement: All electronics below 60°C operating temperature. Target below 50°C under continuous load.

Required: Aluminum heatsinks on each TPA3118, bonded with thermal compound. 60mm fan across the amplifier array. Teensy in a separate thermally‑isolated compartment with passive heatsink. Do not mount the Teensy adjacent to the amplifiers.

Why Quartz?

The target material is not arbitrary. Quartz (SiO₂) has three properties that make it the specified material:

1. Bond angle 143.6° — The Si‑O‑Si bridging angle in alpha‑quartz is 143.6° ± 0.5°, within 0.28% of 144. The fundamental bond geometry of the target material approximates the CTF spatial harmonic.
2. Acoustic wavelength — The speed of sound in granite is approximately 5,950 m/s. At f₀ = 144.069 Hz, the acoustic wavelength is 5950/144.069 = 41.300 m — the resonant cavity dimension.
3. Piezoelectric effect — Quartz converts acoustic forcing at f₀ into electric polarisation at f₀ in the same material. At the 143.6° ≈ 144° geometry, the coupling is self‑reinforcing. The acoustic and electric waves reinforce each other at the same frequency.

The Inscription Connection

After the simulation was built from mathematical first principles, all five parameters from Leedskalnin’s carved inscription mapped onto simulation architecture components:

This mapping was not used to build the simulation. It was discovered afterward. It is hypothesis‑generating — suggesting the inscription may encode calibration parameters for a physical system. It is not historical proof.

Verification Code

import numpy as np
from scipy.integrate import odeint
import math
# Exact f0 from CTF Identity
f0 = 10373 / 72  # 144.069444... Hz
omega0 = 2 * np.pi * f0
Q = 1500.0        # High-Q resonance
alpha = 0.8       # Nonlinear coupling (toy model)
xc = 0.0005       # Threshold displacement
dt = 1 / (20 * f0)
# Verify micro-gap
delta = 10373/72 - 53 * math.e
print(f"f0 = {f0:.9f} Hz")
print(f"53e = {53*math.e:.9f} Hz")
print(f"Micro-gap delta = {delta:.9f} Hz")
# f0 = 144.069444444 Hz
# 53e = 144.068936908 Hz
# delta = 0.000507536 Hz
# 12 emitters, 30° phase increments
phases = np.deg2rad(np.arange(0, 360, 30))
# Amplitude taper (fwd / adjacent / rear-mid / opposed)
taper = np.array([1.20, 1.05, 0.90, 0.80, 1.05, 0.90,
                  0.80, 0.90, 1.05, 0.80, 0.90, 1.05])
def burst_gate(t_idx):
    """9 ON / 27 OFF — 12% duty cycle"""
    cycle = int(t_idx * f0 * dt) % 36
    return 1.0 if cycle < 9 else 0.0
def watch_logic_force(t_val, x, dx):
    """Top-4 coherence selection"""
    contrib = np.array([
        taper[i] * burst_gate(int(t_val / dt)) *
        np.sin(omega0 * t_val + phases[i])
        for i in range(12)
    ])
    scores = contrib * np.sign(dx) if dx != 0 else contrib
    active = np.argsort(scores)[-4:]  # top 4
    force = np.zeros(12)
    force[active] = contrib[active]
    return np.sum(force)
def g_eff(x):
    """Nonlinear coupling — toy model"""
    return 1.0 - alpha * (np.abs(x) > xc)
# Quick validation run (5 seconds)
t_arr = np.arange(0, 5.0, dt)
x, dx = 0.0, 0.0
g_history = []
for i, t_val in enumerate(t_arr[1:], 1):
    F = watch_logic_force(t_val, x, dx)
    ddx = F - (omega0 / Q) * dx - omega0**2 * x
    dx += ddx * dt
    x  += dx  * dt
    g_history.append(g_eff(x))
print(f"Mean g_eff: {np.mean(g_history):.3f}")  # ~0.720
print(f"Min  g_eff: {np.min(g_history):.3f}")   # ~0.180

Conclusions

1. At exact f₀ = 10373/72 Hz, the 12‑emitter dodecahedral array with full architecture produces mean g_eff = 0.720, minimum g_eff = 0.180 in simulation.
2. Performance degrades measurably outside ±0.002 Hz. Sub‑millihertz precision is structurally required — not optional.
3. All four components (watch logic, burst envelope, asymmetric taper, phase bias) are required simultaneously. Geometry alone produces near‑total cancellation.
4. The Teensy 4.0 is the minimum‑cost hardware that provides the 64‑bit FPU precision required to represent f₀ exactly.
5. Quartz‑rich granite is the specified target: Si‑O‑Si angle 143.6° ≈ 144°, acoustic wavelength at f₀ = 41.3 m, piezoelectric self‑reinforcement at f₀.
6. The inscription mapping (15–53–15, 7129/2971, micro‑gap δ, 16th prime) was discovered after simulation construction, not used to build it.
7. A null physical result is explicitly anticipated. The toy model makes no claim about physical reality.

The inscription encodes it. The simulation tests it. The experiment will decide it.

References

1. Gurwell, G. (2026). The CTF Frequency Identity: f₀=(144²+10)/144≈53e. Zenodo Record 19207914.
2. Gurwell, G. (2026). The Leedskalnin Equation. Zenodo. CTF Framework Series.
3. Gurwell, G. (2026). The 144° Bond: Crystal Lattice Structures and Mineralogy. Zenodo.
4. Leedskalnin, E. (1945). Magnetic Current. Rock Gate, Florida.
5. CODATA (2022). Fundamental physical constants. NIST.