-
Notifications
You must be signed in to change notification settings - Fork 1
2026‐04
Caribou Call (First Meeting) — Recording
Attendees: far1no, thomasg, errrks.eth, alperenag, kbmollysuh, zeynepb5793
-
Current State — Frame is built with motors mounted (photo from November). Old battery holders still in the frame from the previous battery design — Julius will cut them out. CAD assembly shown in Fusion with updated battery layout.
-
Architecture Overview
- Batteries insert from the top into the frame; each gets its own battery PCB
- Central signal PCB (similar to Quiver main PCB but without motor power distribution) — receives 12V and 5V from each battery for redundancy
- Battery PCBs: similar to Quiver but higher-rated, with onboard DC-DC converters
- Power redundancy: MOSFET-based 3-source 12V combining (carried over from Feather's power distribution board) will be added to the center PCB
- All remote-controlled via autopilot — no physical pilot toggles (unlike Feather, which was designed with a pilot in mind)
-
Team Availability
- Option 1 (time + energy): Erick, Alex, KBM
- Option 2 (small tasks): Alperen, Zeynep, Thomas
- Julius to reach out individually for time estimates
- Erick: limited availability in April until he moves out of Bay City; will keep up via recordings and GitHub async
-
Meeting Cadence — Weekly on Tuesdays before the Quiver call. Detailed technical sessions can be scheduled separately with fewer people as needed.
-
Testing / V2 Plans
- V1 frame has suboptimal space utilization for the new battery layout — not worth replicating
- Motor beam connectors are welded (labor-intensive) — will be improved in V2
- Thomas / Javelina will build V2 after first test flight: "I'd love to build one"
- Thomas to do US market research / licensing study (self-assigned)
-
Parachute — Never tested on Feather; this frame could be used to gain parachute deployment experience for future builds
-
Tethered Ground Power — Thomas raised it as potentially useful for firefighting. Complex on Caribou (6 independent power units). Consensus: test on Quiver first, then scale up. KBM noted it would require very high-power tethering, likely 3-phase AC-to-DC with serious protection circuitry.
-
Payload Ideas — KBM floated brush bullet dispensing; Thomas more interested in granular or liquid dispensing for Javelina use cases
-
First Flight Target — Julius estimating summer (July-ish). Erick asked about timing relative to "Chicken Festival" (late October) — Julius expects to be flying well before then.
-
Longshot Battery Update — Thomas asked about status. Julius: ready to start; a friend has agreed to work on the BMS. Thomas: "I'm very eager — I'll manufacture and test whatever we need." Julius to set up a call with Ben for onboarding.
-
Quiver Standup (informal) — No Quiver call today. Quick check: no one had open issues. Thomas heading to the workshop.
-
Open Items
- Julius: remove old battery holders from frame; reach out to team individually for time commitments
- Julius: set up onboarding call with Ben (Longshot BMS contributor)
- Thomas: US market research / licensing study for heavy-lift operations
- Thomas: research tethered ground power (talk to firefighting contacts; test on Quiver first)
Caribou Call — Recording
Attendees: far1no, errrks.eth, thomasg, kbmollysuh, alperenag, 21stCenturyAlex
-
Battery Mounting Brackets (KBM)
- KBM started designing slide-in brackets for top-loading batteries
- Material: bent aluminum (no welding)
- Concern about adding holes to the airframe — Julius: 2–4 per battery is fine; reuse existing holes where possible
- Target: prototype drawing by next week
-
Battery Connector PCB (Julius)
- Layout nearly complete; struggling to find the correct 4-pin center connector for the slide-in battery interface (different geometry from Quiver)
- May need to contact Molex directly
-
QuiverHub Port for Caribou (Alex)
- Alex started a Caribou-specific port of QuiverHub — local deployment with Caribou-specific features
- Julius agrees separation makes sense for now
-
Flight Controller — Alex suggested upgrading the FC. Julius: use Quiver's Pix32v6 + FC PCB for first test flight (already designed for swappable FCs). Thomas: good to log hours on Quiver hardware before risking expensive upgrades. Upgrade later.
-
Foldable Motor Arm Connectors (V2 discussion)
- Thomas researched for ~1 hour: downward-folding vertical hinges structurally best (thrust goes into hard stop)
- RJX Hobby sells 50mm side-folding connectors — similar to landing leg adapters; quick option for V2
- Thomas added it back to his to-do list
-
50L Spraying Drone Reference
- Julius chatted with the battery seller — they shared a 50L agricultural spraying drone kit (assembly required)
- Uses Hobbywing motors, CX-13 (~57kg thrust/motor) — 50L payload claim seems ambitious; likely ~50kg MTOW not payload
- Interesting features: weighing modules (3 per arm) for payload measurement, flow meters for liquid dispensing
- Frame is very minimal compared to Caribou
- Caribou still ~4× bigger capacity
- Thomas: weighing modules useful for Quiver too (payload verification, flow rate for liquid/granular dispensing)
-
Thomas — Market Research — Will find time to talk to Kade and continue US market research for heavy-lift operations.
-
Quiver Meeting — Confirmed for tomorrow (not today). No Quiver call this day.
-
Open Items
- KBM: battery mounting bracket prototype drawing (next week)
- Julius: find correct 4-pin connector for battery PCB (contact Molex if needed)
- Thomas: talk to Kade, continue US market research
- Thomas: research foldable motor arm connectors further
- Alex: continue Caribou QuiverHub port
Caribou Call — Recording
Attendees: far1no, errrks.eth, alperenag, kbmollysuh, thomasg
-
Battery Mounting / Connector Sourcing
- Julius reviewed KBM's battery mounting work
- Battery mounts now have four screw points each, mirrored across batteries
- There is enough space left for the PCB enclosure
- Julius found cheaper battery connector options from the original/alternate manufacturers
- Julius prefers a right-angle connector variant
- Some variants have three middle contacts; Julius thinks the extra middle contact can be left unused because spacing appears workable
- Alibaba and extension-cable sources may be viable
- KBM may check local Shenzhen/Huaqiangbei sources and asked Julius for the link/email context
-
Repository / Project Organization
- Julius created an
engineeringfolder in GitHub - Structure includes CAD/build123d and electronics folders
- Each PCB has its own folder
- Goal: each KiCad project should include its own local library for non-standard components
- This avoids Quiver's old issue where symbols/footprints lived only on Julius's machine and makes projects easier for others to open
- Julius created an
-
Battery Connector PCB Architecture
- Julius is designing a battery connector PCB similar to Quiver's but upgraded
- Major change: add an onboard MCU so the battery connector PCB handles precharge, MOSFET/SSR closing, and battery protocol bridging
- Flight controller should only send a kill/shutdown signal in emergencies
- Normal line state has no voltage; pulling the controller pin high shuts down the network
- Proposed MCU: ESP32-C3
- ESP32-C3 provides Wi-Fi/Bluetooth if needed for local parameter checks or a small app and still leaves about 9 spare pins
- Caribou has six batteries, and BMS protocol cannot be assumed to be DroneCAN, so each PCB may need to bridge/convert battery communication to DroneCAN
-
Power Architecture
- Internal supplies: 5 V internal supply and 3.3 V for MCU/control logic
- Internal ground is not isolated from battery negative
- Each battery board will send an isolated 12 V supply to the main PCB
- Julius chose a preliminary 10 A / 120 W per battery target for the 12 V avionics/redundancy supply
- This follows/extends the redundant power supply concept used in later Feather work
-
Current Ratings / PCB Design Targets
- Motor current at ~50% throttle is under ~50 A
- 200 A is treated as peak, not expected continuously
- Julius thinks designing the PCB for 100 A should be sufficient, with proper fuse design
- Main PCB work is next
-
KiCad / Freelancer Workflow
- Julius handed the battery connector PCB task to a new Fiverr freelancer
- Freelancer is expected to send schematics this week and layout about a week after schematic review
- Erick offered to review schematics
- Julius hopes Vector can also help review, especially with KiCad integration tools
- Desired AI/KiCad help includes schematic/PCB review, footprint/rating/pin compatibility checks, design-rule setup consistency, trace widths, layer descriptions, and signal descriptions
- Julius has PCB docs/pinout/architecture files in the repo for review
-
Cabling / CAD Work
- Julius suggested Erick can start cabling work in CAD, especially motor-to-battery wiring
- Erick can import the design into Inventor and leave final PCB connection details flexible
-
Quiver SSR / Lua Script Note
- Julius asked Thomas about flying with the Lua script for closing the SSR
- Julius inspected PCBs Thomas sent: precharge resistors appear overheated/discolored, one has a solder blob, and measured no continuity; likely burned open internally
- The short appears to come from the DC-DC converter
- Julius suspects resistor overheating may have pulled main PCB terminal voltage low, causing the 12 V regulator/DC-DC converter to fail or overheat
- Thomas will install/drop the Lua script on the relevant flight controller next time he sees Boosh/Ben and has emphasized closing the relay as soon as possible
-
Open Items
- Julius: send KBM connector link/email/source context
- KBM: check local connector availability/samples if useful
- Julius: continue main PCB work
- Freelancer: deliver battery connector PCB schematics this week
- Erick: review battery connector PCB schematics when available
- Vector: help review KiCad schematics/PCB when available, starting with inspection before edits
- Erick: start Caribou cabling work in CAD, especially motor-to-battery paths
- Thomas: install/use Lua SSR-closing script on Boosh/Ben flight controller