CBM G1 · Build book
Current STEP-backed mechanical reference · 2026-07-19

Every part retained. Every clearance inspectable.

The CBM G1 engineering package now includes the actual populated Biocoin envelope, upstream-exact dock headers, a full guarded daughterboard, retained cartridge connector, offset battery cradle and harness, board clips, hard stops, gaskets, latches, source materials, and the gates that prevent nominal CAD from becoming a medical claim.

Reader core
Biocoin v1.3Published hardware + firmware
Dry interface
12 @ 2.54Mill-Max 817 at 2.24 mm working height
Wet field
8 × 2×3Six needles per working tile
Nominal fit
12 / 12 passPhysical tests remain explicitly gated
Build logic

Custom spend follows proof.

A passed stage unlocks only the next engineering stage. It never automatically unlocks skin wear.

Gate E

Electrical truth

≤0.5 mV loading error at 1 GΩ; settled cross-talk, temperature, humidity, BLE, and charge checks.

Gate C

Chemistry

Three independent commercial arrays; ladders, residuals, interference, drift, and post-run calibration.

Gate Q

Quality behavior

Every induced fault suppresses the intended output while retaining the labeled raw trace.

Gate M / W0

Mechanical + integrated

Contact, ingress, insertion, fracture, isolation, and full artificial-ISF repeat on the real cartridge.

CP-01

Guarded interface daughterboard

Custom PCB
Frozen specification
EnvelopeØ38 × 0.8 mm full disc
ChannelsNa, K, pH, blank, RE-A, RE-B
Front end6 × LMP7721 + ADS124S08
Upward2×10 + 2×8 Mill-Max 855 dock headers
Downward12-position Mill-Max 817 connector
Release gate: 1.25 mm nominal package edge clearance; ≤0.5 mV loading at 1 GΩ; 6.08 mm Biocoin and 2.24 mm cartridge working heights; reviewed schematic/layout before RFQ.
Manufacturing sequence
  1. Route six guarded islands at radius 15.8 mm and keep each unbuffered run under 8 mm.
  2. Place LMP7721 followers, central ADS124S08, four published dock footprints, and the retained cartridge connector.
  3. Assemble, ionically clean, bake, and verify header/connector coplanarity.
  4. Seat Biocoin in four shell-supported edge clips, compressing J3/J4/J7/J8 by 0.70 mm.
  5. Measure loading, crosstalk, humidity, radio, and charge-state behavior after full mating.
CP-02

Modular polymer microneedle tile

Microfabrication
Frozen specification
Tile3.5 × 5.0 × 0.45 mm target
Array2×3, electrically common
Needles1.00 ±0.05 mm × Ø0.30 ±0.03 mm
Tip≤25 µm radius target
Metal~10 nm Cr + ~150 nm Au; no nickel
Inspection: 100% optical, sampled dimensional/SEM, witness-film thickness, continuity/isolation, locked phantom insertion, compression safety factor ≥3, residual-DMF plan.
Published 2026 starting route
  1. Mix SYLGARD 184 at 9:1; stir 20 min, vacuum-degas 15 min, cure 80 °C for 2 h.
  2. Laser-machine the 2×3 negative mold; 2 W, 100 mm/s, 0.05 mm pitch are starting machine settings, not inspection.
  3. Prepare polystyrene at 300 mg/mL in DMF. Fill mold and centrifuge 5,000 rpm for 5 min.
  4. Dry at 80 °C for 12 h and demold without side-loading tips.
  5. Sputter ~10 nm Cr then ~150 nm Au. Release thickness by witness coupon, not elapsed time.
CP-03

316L metal tile · vendor alternate

RFQ
RFQ requirements
BaseTraceable 316L stainless steel
GeometrySix isolated 30–36G needles, 0.70–1.00 mm
FinishNickel-free gold wet electrode
ReturnsSTEP, 2D drawing, C of C, coating and insertion data
Starting vendorZorayPT ZPT-SMN-001 custom
Incoming QC: dimensions, tip radius, buckling, isolation, coating thickness/pinholes, nickel screen, corrosion in artificial ISF, coating integrity after insertion.
Vendor package must disclose
  1. Material lot and full coating stack, including adhesion layer.
  2. Nominal/minimum coating thickness and uniformity method.
  3. Electrical isolation architecture between tiles.
  4. Cleaning, packaging, residuals, bioburden, and available biological evidence.
  5. Clear nonsterile-research labeling unless a validated sterile package is separately quoted.
CP-04

Twelve-net flex and dry contact boss

Custom flex
Stack + geometry
Core25 µm adhesive-less polyimide
Copper18 µm rolled-annealed target
DiscØ36 inside Ø40 open frame
Tile radius12.0 mm; eight 3.5×5.0 mm landings
Boss2×6 at 2.54 mm; 1.55×1.80 mm pads
Wet-side rule: no nickel, copper, ENIG edge, or conductive adhesive exposed. Three latch lands target ≥17.7 N combined retention.
Assembly sequence
  1. Vendor fabricates and 100% open/short tests the flex.
  2. Converter cuts wet barrier, ARcare passivation, gasket, and adhesive registration features.
  3. Bond separately characterized tiles in the keyed fixture, one chemistry class at a time.
  4. Laminate passivation with ±0.10 mm opening alignment.
  5. Verify continuity, >100 MΩ dry isolation, flatness, pad coplanarity, motion, and ingress.
CP-05

Reference A, reference B, and Pt counter

Custom electrochemical
Reference architecture
RE-APrimary Ag/AgCl measurement reference
RE-BSeparate lane/batch differential sentinel
ReservoirNaCl / PVB protection matrix
CounterCentral Pt six-needle cluster
Truth refeDAQ ET073-1 on the bench
Gate: RE-A/RE-B differential drift ≤1 mV over the intended bench interval; counter remains inside AFE compliance with margin.
2026 starting reference route
  1. Deposit the qualified silver layer on each separate reference tile.
  2. Apply 10 µL of 0.05 M FeCl₃ for 30 s; rinse three times with DI water.
  3. Apply 50 mg NaCl + 79.1 mg PVB in 1 mL methanol.
  4. Dry/condition under the locked process and record offset against ET073-1.
  5. Track each reference's fabrication lane, lot, initial offset, and drift independently.
CP-06 · functionalization

Starting recipes, not invented certainty.

These are transfers from primary methods. Final volumes, thickness, activity, and cure lock only after the G0 area and drift DOE.

Glucose + duplicate

PB / GOx

  1. PB CV: −0.2→+0.4 V, 20 mV/s, 15 cycles in fresh 2.5 mM FeCl₃ + 2.5 mM K₃Fe(CN)₆ + 0.1 M KCl in 0.1 M HCl.
  2. GOx 10 mg/mL mixed 1:1 v/v with 1 wt% chitosan in 1% acetic acid.
  3. Dry, then 1% PEGDE protection. Duplicate uses a separate deposition lane.
β-hydroxybutyrate

HBD / NAD+ / poly-TBO

  1. 1 µL 0.2 mM TBO; dry 45 min; electropolymerize +0.9 V for 6 min in 400 µL 0.1 M PBS.
  2. 1 µL 0.1 mg/mL COOH-MWCNT in 0.5 wt% chitosan / 0.1 M acetic acid.
  3. 1 µL HBD 2.5 mg/mL : NAD+ 40 mM at 1:2 v/v in 0.1 M Tris-HCl; overnight 4 °C.
  4. 1 µL 0.5% chitosan then 1 µL 2.0% PVC.
Sodium + potassium

PEDOT:PSS / ISE

  1. PEDOT:PSS: 0.2 wt% EDOT + 4 wt% PSS; CV 0.2→0.9 V, 50 mV/s, 10 cycles; cure 80 °C 1 h before assembly.
  2. Na: 1 mg ionophore X + 0.55 mg Na-TFPB + 33 mg PVC + 65.45 mg DOS in 660 µL THF.
  3. K: 2 mg valinomycin + 0.55 mg sodium tetraphenylborate + 33 mg PVC + 65.45 mg DOS in 660 µL THF.
pH + blank

PANI / control

  1. PANI: 0.1 M aniline in 0.1 M HCl; CV −0.2→+1.0 V, 100 mV/s, 30 cycles.
  2. The blank matches the physical stack relevant to the validated use while omitting recognition chemistry.
  3. No single blank is allowed to “correct” every analyte without a dedicated experiment.
Method boundary: the dopamine–hyaluronic-acid hydrogel ketone route and a P4VP patent embodiment are separate. G1 does not blend them and call the hybrid published.
CP-07

Shell, passivation, gasket, and adhesive

Print + convert
Mechanical stack
ShellØ44 × 19.0 mm; 1.5 mm wall
BatteryOffset cradle + JST tab harness
AdhesiveSolventum 4077, Ø54 / Ø34 mm
Retention4 board clips + 3 cartridge hooks + 3 PEEK screws
SealsSeparate enclosure-seam and contact gaskets
Release: 0.81 mm battery/component gap, 1.05 mm ceiling gap, 2.31 mm full-Raytac-body keepout, no ingress or insertion-load transfer.
Prototype sequence
  1. Import and center the actual populated Biocoin STEP, not a board envelope.
  2. Print the lower chassis, four clips, hard stops, battery cradle, and top cap in Tough 1500 for fit only.
  3. Install tabbed CP1240 leads through JST ACHR-02V-S; never solder the cell.
  4. Run 500 mate/demate, shell closure, motion, ingress, latch-load, and phantom-insertion tests.
  5. Select production material and sterilization only through the device biological-risk program.
Live sourced BOM

Catalog, open, RFQ, or custom.

Search the engineering CSV. Links point to the manufacturer or primary open-hardware source; prices are intentionally not presented as live quotes.

PhasePart / functionManufacturer / partStatusCritical specificationSource
Loading BOM…
Primary anchors

Methods trace back to sources.

The full evidence ledger is downloadable with the package.

Ketone drift warningSeven-person randomized study with progressive 14-day signal decline.JMIR Diabetes 2026 ↗
Control-heavy human arrayEight cortisol molecular switches plus eight controls; n=3 + n=3.Diabetes 2026 ↗
FDA investigational deviceIDE, NSR, exemption, and IRB decision framework.FDA guidance ↗
Non-negotiable boundary

Exact manufacturing notes are not a self-test protocol.

Bench work stops at artificial ISF, mechanical phantoms, and approved ex-vivo fixtures. A human microneedle study requires IRB review, the correct IDE/NSR or exemption determination, investigational labeling, informed consent, device-level biological evaluation, sterilization and packaging evidence, electrical and battery safety, comparators, monitoring, and adverse-event stop rules. Passing Gate W0 does not waive any of that.