I've now thoroughly reviewed both PDFs. Let me give you a clear summary.
Yes, both USB charging and battery protection are already covered in your battery architecture — they're part of the self-contained battery pack design you've been developing. Here's the breakdown:
The battery pack includes a full USB-C PD charging system built into the electronics bay at the bottom of the scaffold: ppl-ai-file-upload.s3.amazonaws
- USB-C panel-mount port (IP67, on the side of the battery end cap)
- PD trigger board (ZY12PDN or STUSB4531, ~31×15×4mm) — negotiates 20V from any charger 45W–240W automatically
- Boost converter (XL6019, 20V→42V CC/CV) — steps up to pack charge voltage
- Total electronics fit in ~60×45×25mm, tucked into the electronics bay
The key design point: all electronics live on the battery, not in the frame. You pull the battery out, plug your MacBook charger into the USB-C port on the side, charge on the kitchen table. The frame side is 100% passive — just flat copper pads. ppl-ai-file-upload.s3.amazonaws
A 10S BMS is included in the electronics bay, handling: ppl-ai-file-upload.s3.amazonaws
- Overcurrent, overvoltage, undervoltage protection
- Short-circuit and temperature cutoff
- Cell balancing during charge
- Per-cell fusing was deliberately skipped — at 3.5A per cell in 2P, the BMS alone handles everything adequately
The connector design settled on an elevated pad island + recessed pogo pin dock approach: ppl-ai-file-upload.s3.amazonaws
- Frame side: Elevated flat gold pads on a raised "speed bump" — water sheets off, no pooling
- Battery side: 6-pin pogo block recessed deep inside a dock cavity — finger-safe by geometry (IP2X-style)
- 2 power pins + signal/thermistor/comm pins
Your notes confirm the DT dimensions are not locked down yet, and you've been exploring both round and oval options: ppl-ai-file-upload.s3.amazonaws
| Discussed Size | Context |
|---|---|
| 50mm round | Original spec for 2×2 grid of 21700 cells |
| 55–60mm | Referenced for fitting 21700s with BMS/wiring room |
| 43×59mm oval | The 1-2-1 diamond layout (narrower, taller) |
| ~50×65 (your current thinking) | Wider than the diamond, gives more lateral room |
The 50×65 oval gives you more cross-sectional area than the 43×59 diamond and better lateral stiffness for pedaling loads. The tall axis (65mm) resists braking/vertical loads while the narrow axis (50mm) keeps the frame slim between your knees. Since you're building in aluminum (not steel), you have freedom to specify custom oval tubing from the fabricator. ppl-ai-file-upload.s3.amazonaws
The scaffold cage parameters (currently cagew = 53, cageh = 48 in the OpenSCAD file) would need updating to match whatever oval you settle on. With a 50×65 outer and ~2mm wall, you'd have roughly 46×61mm inner bore — plenty of room for the 2×2 cell grid (~43×43mm footprint) plus wiring channel and air gaps. ppl-ai-file-upload.s3.amazonaws
Bottom line: your electronics stack is already specced and costed at $22–38 total for BMS + USB-C PD + boost converter. The DT oval just needs to be finalized around whatever gives comfortable clearance for the scaffold + 1mm air gap on each side. ppl-ai-file-upload.s3.amazonaws
You're right — I was wrong about the path. Let me reconsider.
What's Actually Inside the Steerer
The carbon fork steerer tube is not a clear conduit. Inside it you have:
The brake hose goes in the space between the steerer outer wall and the head tube inner wall — not inside the steerer itself. That's the cable routing channel you're describing. [ppl-ai-file-upload.s3.amazonaws](https://ppl-ai-file-upload.s3.amazonaws.com/web/direct-files/attachments/139085844/51f251cf-3b46-47fd-9df4-4b92d46f5b41/frame-build.pdf)
So Can the Wire Get Out?
Actually yes — it just doesn't go through the steerer. The antenna wire exits at the edge of the top cap, not down the center:
The coax (~1.3mm diameter) just sneaks out at the seam between the top cap edge and the top spacer. It's the same gap where water already gets in — it's not sealed. Then it drops into the same routing channel as your brake hose and front light wire, between the steerer and head tube wall. [ppl-ai-file-upload.s3.amazonaws](https://ppl-ai-file-upload.s3.amazonaws.com/web/direct-files/attachments/139085844/51f251cf-3b46-47fd-9df4-4b92d46f5b41/frame-build.pdf)
Is That Clean Enough?
Honestly — it's a visible wire running from the top cap edge down into the headset area. It's thin, and you could color-match it black, but it's not invisible. You'd see a hair-thin black wire at the stem.
If that bothers you, the bottle cage bolt on the seat tube might still be the cleaner option despite the manufacturing challenge — or you could simply accept that BLE range through aluminium will be reduced but not zero, and skip the external antenna entirely for v1. How critical is the GPS/BLE tracking for the first prototype?