Ashley Kalkowski

DESTROYING BOULDERS / ENGINEERING EDITION E01

Inside
the exciter.

The housing is only the beginning. Follow the driver, the moving parts, the interfaces and the energy through the next generation of the research.

Research direction: Ashley Kalkowski.
06 October 2026, 19:23 UTC · Digital assemblies and numerical experiments.

Take the assembly apart
Original CAD rendering of the E01 component-based acoustic assembly
ACTUAL CAD GEOMETRY / UNBUILT ASSEMBLY / E01
2Digital component assemblies
109Placed CAD components
52Original + correction integrations
0Physical tests

NEW / E02 + EL01 + M01

Continue beyond this assembly.

Electrical integration

Inspect the selected candidate supply, USB audio adapter, matched panel kit, connector map, measurement accessories and remaining qualification gaps.

Open the electronics & wiring edition

Reset and braking

112 conditional records quantify return demands and retain the pickup failures. Powered repetition is still an unbuilt mechanism.

Open the reset study

The specimen volume

A separate 3D finite-element material comparison adds calculated motion, stress and a video replay, with unresolved convergence exposed.

Open the 3D material lab

This E01 page preserves the original assembly/model edition. EL01 supersedes its pending supply/audio selections and uses the complete KAB-PMV3 panel kit; KAB-FC remains a partial alternative, not an additional required purchase. The updated electronics module is external to the acoustic cavity.

01 / A MORE COMPLETE QUESTION

What makes it operate?

The acoustic route converts electrical input into diaphragm motion. The impact route releases a known moving assembly and accounts for its contact, rebound and capture. Both now have explicit parts and model boundaries.

A / AIRBORNE

A real motor and suspension.

Two actual driver sizes have separate manufacturer parameters. Rear-chamber volume changes stiffness and the response. Current, back EMF, motion and power are calculated together.

I / CONTACT

A captive impact reference.

Metal guides, a release pin, a replaceable tip and a secondary capture define one gravity-driven event. Repeated powered timing is a subsequent mechanism requirement.

E / EVIDENCE

Expose the assumptions.

Kernel geometry checks, exported parts, numerical energy ledgers and failed cases remain inspectable. None of these is a physical fracture test or a promise of a chosen crack.

02 / EXPLORE THE ACTUAL PARTS

See how it fits.

The viewer uses tessellated solids from the CAD export. Select a part, separate the assembly or inspect the section and dimensioned images below.

DIGITAL ASSEMBLY / MILLIMETRE GEOMETRY

Loading assembly geometry…

Reference component geometry uses the available catalog datums. Unknown terminal envelopes, part tolerances, material behavior and purchased revisions remain explicit fit gates. A valid CAD solid does not qualify a manufactured part.

03 / THE ACOUSTIC OPERATING MODEL

Change the inputs.
Watch the response.

This is a small-signal electrical and mechanical model at the driver terminals. The graph is calculated locally from the documented parameters. It does not generate sound or control equipment.

Exploration limits are software bounds, not an approved operating range. Net volume excludes the driver, bosses and other occupied space.

Input voltage means actual differential RMS voltage at the speaker terminals. Computer volume and digital audio amplitude are not calibrated voltage. The model includes the catalog effective moving mass and aggregate mechanical loss; neither is counted again as a separate radiation term. Box pressure is internal pressure, not sound pressure at a rock.

Equations, domain and energy boundary

The reciprocal model couples coil voltage and current to diaphragm velocity through the force factor. Suspension stiffness and ideal sealed-air stiffness act in parallel. In sinusoidal steady state, real electrical input equals copper loss plus the catalog aggregate mechanical loss. The latter does not independently identify radiated acoustic power.

L di/dt = V − Re i − Bl v
M dv/dt = Bl i − Rms v − K x
dx/dt = v
K = 1/Cms + ρ c² Sd²/Vbox

Temperature drift, nonlinear suspension and motor behavior, cone breakup, leaks, enclosure wall compliance, field focusing, specimen coupling and fracture are outside this model. The selected 20–1000 Hz range is a numerical exploration, not a validated band for every omitted effect.

04 / THE IMPACT OPERATING MODEL

An impact does not
end at first contact.

D02 found every saved striker still moving at the end of its earlier trace. E01 adds gravity, guide loss, a target and a compliant secondary capture, and retains motion and failures through the end of each new run.

The eight fine-resolution cases retain 1 geometric-boundary failure(s). 2 of 8 meet the prescribed numerical near-rest criterion over the final 50 ms. These are results for the declared hypothetical contact and buffer coefficients, not measured capture ratings.

CaseTarget peakBuffer peakRetained endTravel outcomeMotion outcome
I00 · Nominal target and guide28.81 N0 N0.6 sCompleted windowRest criterion not met
I01 · Maximum permitted mass/drop37.88 N0 N0.6 sCompleted windowRest criterion not met
I02 · Nominal missing-target capture0 N38.81 N0.6 sCompleted windowNear rest in final50ms
I03 · Maximum missing-target capture0 N53.61 N0.6 sCompleted windowNear rest in final50ms
I04 · Deliberately soft undamped capture0 N3.004 N0.1417 sBoundary crossedRest criterion not met
I05 · Stiffer missing-target capture0 N55.68 N0.6 sCompleted windowRest criterion not met
I06 · Higher smooth guide loss20.57 N0 N0.6 sCompleted windowRest criterion not met
I07 · Softer target contact16.18 N0 N0.6 sCompleted windowRest criterion not met

The nominal dynamics case assumes a 12 g moving assembly released 85 mm above the target. The CAD material estimate is 12.2047 g; neither value is a measured build mass. Acceptance requires weighing every moving part. The 20 g upper mass and 95 mm maximum release-to-hard-boundary descent correspond to 18.632635 mJ of ideal gravitational work. Guide, contact and capture behavior are assumptions until measured. The manual cartridge establishes one impact; powered reset and controlled repetitive excitation are not implemented here.

05 / THE COMPLETE BENCH CHAIN

Connect the interfaces.

The printed head is one part of the system. The amplifier, supply, cables, measurements and specimen fixture belong in the design and the eventual cost.

  1. PC & wired audioDefined digital input. Gain and timing require measurement.
  2. External amplifierPurchased board, matched harnesses, enclosed low-voltage supply.
  3. Driver terminalsDifferential voltage and current define the model input boundary.
  4. Driver & headMechanical motion, chamber, mounting and leakage.
  5. Specimen & sensorsDelivered field and target response still require calibration.

Fasteners and fit

Printed structures, metal guide and wear surfaces, purchased drivers and fasteners have different jobs. Use the dimensioned parts and fit coupons; do not scale the complete assembly to change driver size.

Assembly without guessing

The component record identifies included harnesses and unresolved terminal fits. A solder-free build is an objective; it is not established until the actual connector sizes and strain relief are checked.

Cost and scale

The earlier under $1,000 first-module target remains a constraint. Stock, delivered cost and a complete calibrated bench total have not been verified. These small assemblies do not establish boulder-scale delivery or controlled fracture.

SubsystemCandidate / custom partCurrent evidenceManufacturing or calibration gate
Acoustic driverDayton ND64-4; ND91-4 is a separate modelCatalog datums, own T/S parametersTerminal dimensions and actual revision
Housing, baffle, grille and gasketsOriginal custom CADPer-part STEP and custom STLProcess, material, gasket compression and leakage
Fasteners and guidesNominal metal component referencesExplicit holes and assembly stackThread/grade, engagement, tolerances and moving mass
Amplifier and wired harnessDayton KAB-215v2 with KAB-FC AUX kitManufacturer interface documentsGain, clipping, calibration and exact connectors
External supply and PC audio interfaceParts selection pendingInterface requirements documentedModel, complete cost and measurement compatibility
Capture buffer and target fixtureHypothetical compliance in numerical studyFailure cases retainedMeasured stiffness, damping, strength and mounting

Electrical interfaces and characterization sequence · Manufacturer source register

06 / DOWNLOAD & REPRODUCE

The assembly.
The code. The evidence.

Original parametric source, STEP assemblies, individual parts, executable numerical models, frozen inputs and results. Earlier versions and negative outcomes remain available.

Downloads use lossless .tar.xz packaging. Extract each into a fresh folder with Python 3: python -m tarfile -e bundle.tar.xz new-folder. Every extracted member was checked against the original verified source ZIP. Frozen source notes may still name that original ZIP; the commands after extraction and all member bytes are unchanged. The hashes shown here identify the new download containers. Complete bundle contents and file hashes. Historical numerical traces remain individual downloads below.

Component assemblies

Parametric CAD source, real STEP exports, custom-part meshes, assembly views, dimensions and component/fit records.

Download · 3.10 MB

SHA-256
412501d77644cd84ff82aff70aa14066b0ce99d4ff91ecfdf5f494b03de83e52

Executable operating models

Frozen inputs, Python models, tests, results, numerical figures and reproduction commands. Full retained NPZ histories are separate.

Download · 1.34 MB

SHA-256
bb3074fbb9b15d000fa2175b9bf6d1f53fb2be4f5214f378c0be76bd16ed9252

Independent review and component evidence

Independent numerical audit, manufacturer parameter records, electrical interfaces and a blank measurement record.

Download · 0.19 MB

SHA-256
94b0b8de0f3a7aee3aed40cfc6d33ae13489e0fee7e5459989f95dd65292e78c

Original matrix:972 algebraic frequency-response points, 24 acoustic ODE integrations and 16 impact integrations, completed in 137.11 seconds.

Geometry correction:A single-solid check rejected four mounting bosses that only touched the enclosure wall. Joining them changed the nominal air volume from 0.706091 to 0.703978 litres. The separately frozen addendum recomputed 486 affected algebraic points and 12 acoustic integrations in 36.09 seconds. The current 40-integration composite inherits 28 unchanged results; all 52 historical integrations remain available.

Current numerical outcome:Acoustic numerical screens did not all pass; coarse/fine refinement screens passed. One corrected coarse 20 Hz acoustic trace has returned-energy samples as low as −3.10239×10⁻¹² J, failing the frozen −1×10⁻¹² J nonnegative-ledger tolerance. Its finer counterpart passes that gate. The failure remains recorded; accurate amplitude, phase and overall energy balance do not erase it.

Independent sampling finding:A stricter review also detected tiny decreases in sampled cumulative-energy ledgers in 14 historical acoustic traces, up to 1.34×10⁻⁸ J, or 4.42×10⁻⁷ of positive supplied energy. Dense-output interpolation is the likely cause; the original strict audit and frozen results remain intact. Audit agreement means a saved outcome was reproduced, including failures.

Current composite summary · Frozen correction plan · Original matrix · Author tests

Independent review files and geometric acceptance records are included in the downloads. A mathematical check, a valid solid and a physical test answer different questions.

Exact E01 public-file manifest · D02 demands and C02 contact-path correction · Earlier C01 concept family

07 / WHAT WOULD MAKE THIS BETTER?

Measure the missing interfaces.

Build and characterize one head

Measure fit, mass, impedance, terminal voltage/current and motion. Compare measured resonant behavior with the retained predictions without silently changing the original model.

Qualify capture and repetition

Measure guide friction and buffer behavior, then design a powered reset mechanism with its own timing, thermal and failure analysis. A reserved mounting interface is not a working reset drive.

Connect to a specimen model

Identify delivered force or pressure and specimen response before adding calibrated fracture mechanics. The route to a chosen split needs crack-path evidence, not just a resonance peak.

Research direction: Ashley Kalkowski. Computational implementation and review used Codex assistance. Source provenance, established mechanics and third-party dependencies remain attributed. This release does not assert patentability, exclusive rights, manufacturer endorsement or a new blanket reuse licence.