# Inherited evidence - R3.3 harness boundary

This document preserves the R3.3 bounded connector/adapter study as source context for website 28.1 / engineering R3.4. It is not a new R3.4 harness solve, installation drawing or fabrication release. R3.4 changed printed geometry; the complete harness remains unvalidated. Use the current Engineering Handoff and Assembly_and_Nonprinted_Parts.md for the current print set and assembly boundaries. Historical file names below identify the underlying study, not additional files to print or orders to place.

The original source follows unchanged.

---

# R3.3 bounded harness assessment and wiring schedule

## What is and is not delivered

This is a finite socket/exit/adapter assessment and corrected terminal reference. It is not a restarted full harness solve, a harness installation drawing or an electrical release. The checked R3.3 mechanics remain fixed. Local bend probes and terminal spheres in the reference STEP must not be fabricated or mistaken for complete conductor routes.

All measured overlaps below refer to the adopted reference geometry against the actual fixed R3.2 A components and checked R3.3 shells/keepers. CK and stock PH female housings are documented-size **conservative exterior reservations**, not detailed supplier STEP solids. The additional PH male adapter uses an unchanged copy of the actual B2B-PH header model. Intentional contact with its own mating header is excluded from CK clearance checks.

## CK proposal, not approval

JST CK is a compatible cable-side family for PH through-hole top-entry headers. The documented mounted height is **6.9 mm**, socket body height **5.15 mm**, body depth **4.2 mm**, and mounted width **4.8 mm**. Two-position body width is **6.0 mm**; five-position width is **12.0 mm**. ([JST CK catalog](https://www.jst-mfg.com/product/pdf/eng/eCK.pdf))

| Interface | Normal / reverse candidate | State |
|---|---|---|
| J2 battery board interface | 02CK-6H-P / 02CK-6H-PC ([JST](https://www.jst-mfg.com/product/pdf/eng/eCK.pdf)) | Proposed pigtail only; stock battery PH plug cannot simply be replaced |
| J10 speaker board interface | 02CK-6H-P / 02CK-6H-PC ([JST](https://www.jst-mfg.com/product/pdf/eng/eCK.pdf)) | Alternative only; stock UL1007 tails are not assumed IDC-compatible |
| J3 microphone board interface | 05CK-6H-P / 05CK-6H-PC ([JST](https://www.jst-mfg.com/product/pdf/eng/eCK.pdf)) | Proposed; physical pin order remains unassigned |

These normal/reverse models have the same documented exterior envelope but reverse the circuit-1 end. The geometric result is therefore shared; normal/reverse is not permission to guess polarity or pin order. ([JST CK catalog](https://www.jst-mfg.com/product/pdf/eng/eCK.pdf))

CK permits **UL1061 or UL10272, AWG26/28, seven-strand tinned annealed copper, insulation OD 0.75–1.05 mm**. The quoted 2 A rating is at AWG26. Current, transient capability, derating, IDC tool/process, pull retention and polarity remain unqualified; the lower-rated KR family was not substituted. ([JST CK catalog](https://www.jst-mfg.com/product/pdf/eng/eCK.pdf))

The local wire-exit study uses 1.05 mm OD, two estimated exit-Z cases and concentric quarter bends, with a 2.5 mm minimum bend radius and preserved 2 mm conductor spacing. JST does not supply a separate exit-centerline or minimum bend-radius datum in this catalog, so these remain adjustable installation assumptions. Local exit probes are tested individually as well as for inter-wire interference; they are not full service loops.

## Concrete local limits

| Socket reservation | Rear-material overlap | Retained antenna overlap | Minimum rear gap if clear |
|---|---:|---:|---:|
| J2 | 2.382282 mm³ | 4.199996 mm³ | 0 mm |
| J10 | 0 mm³ | 0 mm³ | 0.340 mm |
| J3 | 0 mm³ | 0 mm³ | 0.260 mm |

J2 also overlaps the retained dielectric seat by **0.128310 mm³** and is only **0.200 mm** from the lower antenna frame. The full local shell geometry, rather than a flat rear-wall approximation, controls these results. Do not mill, shave or compress the antenna/seat or force the rear cover over this reservation.

J3/J10 outline clearance is not wire-installation approval. All 12 tested local concentric-bend configurations have reservation conflicts with surrounding parts, although their conductors do not intersect each other. Individual cases, actual radii, exit coordinates, collisions and sub-0.7 mm clearances are in `harness_ck_adapter_audit.json`. These results reject the tested bend patterns; they do not prove that every alternative routing pattern is impossible.

## Stock battery remains intact

The proposed adapter chain is:

`stock Adafruit pack and leads -> stock PH female plug -> additional PH male header -> insulated permitted-wire pigtail -> proposed CK socket -> unchanged J2`

No two female sockets are placed on one male mating face. No stock battery connector is cut off or reterminated. The added male-header model is copied rigidly from the baseline purchased B2B-PH geometry; it is an additional proposed part, not a replacement board header.

Two candidate virtual mounting datums were evaluated: **(−22,+3,−6.25)** and **(−22,−3,−6.25) mm**. The model includes the stock PH female exterior, intact male tails, local pigtails, separate solder/insulation allowances and plug-only −X withdrawal samples at 0/2/4/6 mm. Both poses conflict with the front shell. Per-member overlap values are reported without summing overlapping male/female mating reservations. No adapter cradle or attachment process is released.

The main boundary reference shows the −Y candidate only for review, not because it passed. Separate rejected-pose STEP files preserve both alternatives. The +Y pose has less measured conflict: its exact male-header model overlaps the front shell by 3.477528 mm³ and its stock-female reservation by 13.688531 mm³; the −Y values are 5.017969 and 30.943000 mm³ respectively. These per-member volumes must not be summed. Full stock-lead dressing, installed connector seating, strain relief, unplug access and cover-opening slack remain unresolved. The nominal 100 mm battery leads are not treated as a measured length or a guarantee that a future adapter route will fit.

## Correct physical endpoints and electrical hold points

| Branch | Correct CAD location / retained interface | Electrical and mechanical state |
|---|---|---|
| Battery | Existing lead-cap datums (−30,47.28,−9.36) and (−30,45.28,−9.36); raised region X−28…−16, Y43.78…48.78 | J2 pin1 VBAT_CELL, pin2 GND from owner release. Pack maximum, actual lead length/exit and stock polarity require sample checks |
| Speaker | Actual speaker spans Y15.10…68.90; supplied model has frame/magnet solids but **no wire-terminal solids** | No exact solder endpoint is invented. J10 is SPK_P/SPK_N, neither output ground; received lead exit must be located |
| Microphone | Six actual Ø1 mm holes at X−6.35,−3.81,−1.27,1.27,3.81,6.35; Y−9.428; rear board face Z−4.85 | Five-wire J3 net set retained. Pad IDs in CAD are location markers, not an approved net map |
| LED | Four free lead ends at X−14.6, Y2; Z−5.095,−6.365,−7.635,−8.905 | Geometric order is not automatically lamp pin numbering. Confirm physical orientation before termination |
| PTT | Four actual nominal terminal free ends: X−29.3, Y13.75/20.25, Z−11.45/−6.95 | Two conductors needed after continuity identifies the correct contact pair; input-to-J9 assignment remains open |
| Volume up | X30.4, Y19.75/26.25, Z−11.45/−6.95 | Same contact-pair and input-mapping hold |
| Volume down | X30.4, Y−9.25/−2.75, Z−11.45/−6.95 | Same hold |
| Front play | X−6.25/+6.25, Y−49.5/−44.5, Z−10.6 | Same hold |
| Cellular | Existing J5 CELL_RF and R3.2 A 150 mm route, 141/159 alternatives | Untouched. No GNSS or second cable; RF spacing remains open |

Four physical controls require four signal branches **and four local common-return branches** leading to the common J9 ground. A five-wire trunk alone is not a completed four-switch harness. No common-return splice location or complete branch routing is approved in this boundary package. J12 stays reserved.

### Released mappings versus proposals

J8 is pin1 red, pin2 green, pin3 blue, pin4 common-cathode GND. The lamp mapping is J8-1 → lamp1, J8-2 → lamp4, J8-3 → lamp3, J8-4 → lamp2. The supplier requires first lead forming at least 3 mm from the lens base, no more than two bends, and no forming after mounting; insulation and solder space must preserve these constraints. ([Kingbright drawing](https://www.kingbrightusa.com/images/catalog/spec/WP154A4SUREQBFZGW.pdf))

Adafruit 3421 SEL low selects the left channel and high the right channel. A local SEL-to-GND link is only a proposed left-channel implementation; no physical pad or released J3 assignment has been invented here. Keep the bottom-port orientation and seal unchanged. ([Adafruit pinout guide](https://learn.adafruit.com/adafruit-i2s-mems-microphone-breakout/pinouts))

The selected speaker has 152 mm UL1007 28 AWG bare leads. Those leads are not within CK's listed UL wire styles; a CK implementation needs a qualified permitted-wire pigtail/joint or another separately approved termination. Do not silently IDC-terminate them. ([Same Sky drawing](https://www.sameskydevices.com/product/resource/cms-40504n-l152.pdf), [JST CK catalog](https://www.jst-mfg.com/product/pdf/eng/eCK.pdf))

For crimped PH connections, proposed PHR-2/4/5 housings with SPH-002T-P0.5S contacts accept AWG30–24 and 0.8–1.5 mm insulation OD. This is a compatible proposal, not an approved harness BOM. ([JST PH catalog](https://www.jst-mfg.com/product/pdf/eng/ePH.pdf))

## Concrete fallback requiring approval

The documented board-header options remain S2B-PH-K-S(LF)(SN) for J2/J10 and S5B-PH-K-S(LF)(SN) for J3. Side entry reduces the normal mated height to 4.8 mm but creates a 9.6 mm lateral envelope. No header was changed. Pitch/drill overlay, pin1/orientation, populated-board and antenna/battery clearance, rework suitability, electrical current and physical mating must all be checked before authorizing replacement. ([JST PH catalog](https://www.jst-mfg.com/product/pdf/eng/ePH.pdf))

The bounded study establishes failures of the tested reservations and adapter poses, not a mathematical proof that every possible harness architecture is impossible. A full closed-case harness is still unvalidated. Preserve the mechanical checkpoint and resolve the J2/adapter architecture explicitly before another routing revision.