2026‐07

July 7, 2026

Caribou Call — Recording · HearHear transcript

Attendees: far1no, kbmollysuh, errrks.eth, alperenag, thomasg, 21stcenturycotyk

• PT2 Brainstorming Kickoff

  • KBM is already writing PT2 ideas for the forum. Text plus rough hand drawings is acceptable for now; CAD is optional.
  • Julius wants ideas comparable near the end of July and will create a thread to collect them.
  • Nils (off-Discord) is also working on his own PT2 CAD version; Julius asked him because he did the first frame.
  • KBM flagged that the current design has limited interior/cockpit space, so a practical-flight configuration likely means an interior redesign — a main motivation for collecting different-design ideas.

• Layout Ideas (ideas only, nothing planned)

  • KBM prefers staying hexacopter but suggested adding forward-thrust propulsion so the large airframe doesn't tilt in forward flight (efficiency loss from tilting/turning a big copter).
  • Alperen referenced the Sikorsky S-97 Raider concept — the pusher-propeller aspect only, not coaxial main rotors.
  • Julius: interesting, but depends on mission profile.
  • KBM is also curious about asymmetric/non-circular propeller layouts (Joby-style rows/columns); he recalls PX4 supports these and will research.
  • Firmware direction: stick with ArduPilot for the next version unless someone makes a concrete case to switch.

• Mission Profile / Agricultural Ops (Thomas)

  • Main Caribou target: dispensing granular agricultural product, payloads maybe up to 1,000 lbs.
  • Ranchers/landowners vary between blanket treatment of whole properties and precision work. Framing: Quiver for precise spot treatment avoiding certain plants; Caribou for large areas of catclaw/undesirable species. Some jobs discussed run into thousands of acres.
  • Kate hasn't used a large DJI drone, but Thomas has seen comparable operations: a big cargo trailer running generators as a pit stop for battery swaps and product reload — likely the operating pattern for Caribou.
  • Julius asked Thomas to get a satellite image of a helicopter-treated area from Kate to gauge typical job size, battery-change frequency, and turnaround patterns; Thomas will ask.

• PCB / Assembly Status (Julius)

  • C90D PCBs arrived: connectors and screw terminals fit, the deliberately tight fit works, no footprint issues. Photo to follow for KBM.
  • CMAIN is with DHL; CBC PCBs still in manufacturing, expected to ship around end of the week.
  • Assembly progressing: motor orientations fixed, screws re-tightened/exchanged for better-fitting ones, one ESC mounted and fitting well.
  • Main enclosure printing soon. CBC enclosure change agreed: Julius adds a MOSFET heatsink (possibly larger than Quiver's) to the CAD; KBM cuts the enclosure opening. No fan — Julius expects airflow to be sufficient.

• Test Timeline

  • Target: start testing the week of July 20, leaving a ~2-week window before Phase 1 closes at the end of July.
  • First flight mission: short hover, land, check everything, then small incremental inputs.
  • Baseline GPS without RTK (Julius previously used an RTK service; no base station on hand).

• Other

  • Alex will be in China for about a month from July 16 (ecosystem-dev trip), visiting KBM before KBM's own travel at the end of July.

• Open Items

  • Julius: create and seed the PT2 brainstorming thread.
  • KBM: post PT2 ideas (text + drawings) before end of July; research asymmetric prop layouts / PX4 support.
  • Thomas: ask Kate for a satellite image of a helicopter-treated area.
  • Julius: send KBM a photo of the C90D fit.
  • Julius: track CMAIN/CBC deliveries; continue assembly prep; print main enclosure.
  • Julius/KBM: heatsink model onto CBC CAD → enclosure opening cut.
  • Team: aim for test start the week of July 20.

July 14, 2026

Caribou Call — Recording · HearHear transcript

Attendees: far1no, kbmollysuh, thomasg, errrks.eth — kept deliberately short (Julius outside; audio rough in places)

• PT1 Hardware Status (Julius)

  • CMAIN boards have arrived (were with DHL as of July 7).
  • All enclosures are printed, including the battery-pack ones.
  • CBC (battery connector) PCBs arrive July 15–16.
  • The CBC enclosure has already been adjusted with the heatsink opening discussed on July 7; KBM will review it (he owns the heatsink cut).
  • Next step once CBC boards land: work toward a first motor test, then reassess.
  • KBM asked for detailed assembly photos of the PCBs + printed enclosures; Julius will post them.

• PT2 Brainstorming (KBM)

  • KBM's ideas + a rough concept sketch are now in his forum post; he invited Julius to mark items doable / not doable.
  • Bracing (strut/wire triangulation, like old braced airframes): strengthens structure but only suits slow aircraft — KBM thinks it's viable for PT2; Julius wants to check whether it actually buys an advantage.
  • KBM continues research on PEEK structural parts.

• Higher Voltage

  • KBM recommends an eventual move to higher pack voltage. Julius: it's a cost question — Hobbywing X15 sets are cheap as complete COTS units at their designed voltage; going beyond means custom propulsion and Feather-style cost growth. Battery-architecture changes could reopen this later; parked for now.

• Ag-Mission Economics

  • The simulation with Alex's tool plus the job info from Thomas suggests a few real missions reach break-even on the drone cost quickly.
  • Thomas agrees with caveats: these jobs are already flown by helicopters/fixed-wing today, so demand exists — but jobs must actually be won, and chemical cost is a significant subtraction from revenue.

• Texas Build

  • Thomas wants to build a Caribou at his Texas workshop; his new workshop helper is a fairly skilled welder. Julius gives the thumbs-up when the design is ready to send.

• Open Items

  • Julius: post detailed PCB + enclosure assembly photos.
  • Julius: receive CBC PCBs (Jul 15–16); proceed toward first motor test.
  • Julius: mark KBM's PT2 forum ideas doable/not doable; evaluate bracing.
  • KBM: review the adjusted CBC enclosure (heatsink opening).
  • KBM: continue PEEK structural parts research.
  • Thomas: await Julius's go-ahead for a Texas Caribou build.

July 21, 2026

Caribou Call — Recording · HearHear transcript

Attendees: far1no, kbmollysuh, errrks.eth, thomasg

• PT1 Assembly and Electronics

  • The CBC PCB and its enclosure are working. Two batteries are connected and a first single-motor spin test was recorded and posted; the CAN connection is up.
  • The CBC firmware translates the Tattu battery protocol into DroneCAN and nodes are receiving IDs; tests look fine.
  • Next: finish battery and ESC wiring, then a full motor test, then mount GPS and camera by the end of the week. Trailer and tethering preparation follows the week after.

• First Power Run

  • No extra copper pads are needed for the C90D's large connection.
  • The first run stays conservative — roughly 50% throttle, ramped slowly with temperatures watched. KBM remains wary of the high-current contacts.

• Battery Telemetry Architecture

  • Julius routes Tattu telemetry through the ESP32 on the CBC and converts it to DroneCAN, the same role the Tattu bridge plays on Quiver's Raspberry Pi. He built it in because he found no option to switch his 4.0 packs to native DroneCAN in the app.
  • Erick reported the opposite result on Quiver: Tattu 4.0 telemetry appeared directly on the CAN bus in the optional-hardware device scan and took a node ID without manual assignment.
  • Julius will try connecting the battery CAN directly to the main CAN to see whether the bridge can eventually be removed. The bridge also covers non-4.0 packs, where native support does not exist.

• Market Economics

  • A West Texas helicopter pilot quoted $15–18 per acre for granular dispensing, helicopter cost only and excluding chemical — favourable framing for Caribou in that market.
  • Method context: roughly 10 lb per acre of granular product, with helicopters carrying a few hundred pounds, flying for hours, and hovering over a trailer to refill without landing.
  • Endurance is the drone's weak spot against that. A hybrid turbo-generator Caribou was raised as a long-term idea worth reflecting in phase 2 planning.

• JLC USB Connector Defect

  • The USB connector for the ESP32 came back from JLC with pins effectively floating — likely mis-seated or lifted during reflow, and the thick copper makes hand-soldering nearly impossible.
  • With 10 boards in hand, affected units are set aside. Options are wiring cables directly or desoldering and replacing with low-temperature solder paste.

• Tethering Rig

  • Concept: the drone sits on the car trailer over a pulley, with the rope anchored to the car or trailer and a second rope used to keep the main rope taut, tended by a person. Working radius around 5 m.
  • Automatic tensioners were rejected as expensive and not clearly useful; a second person adjusts better. KBM's rubber-band or partial spiral-cable idea is held as plan B — Julius's concern is that a slack rope gets pushed around by propwash.
  • Julius will post a drawing of the rig for feedback.

• Safety Kit

  • Strobe lights are ordered and mounted on 3D-printed mounts using existing screw holes directly under the motors; very bright and well placed.
  • KBM will post safety and reflective sticker suggestions for Julius to order.
  • Open concern: the printed landing feet are infilled rather than solid, with only ~2 cm of printed section under the tube. A collapse would tilt the aircraft into a possible prop strike.

• Availability

  • KBM is in his final week in China, then 11 days in Japan — reachable by message daily but tablet-only, with Onshape available for emergencies and no Fusion. The main design is essentially complete.

• Open Items

  • Julius: finish battery and ESC wiring, then run the full motor test.
  • Julius: mount GPS and camera by end of week; trailer and tethering the following week.
  • Julius: test battery CAN directly on the main CAN to see if the ESP32 bridge can be dropped.
  • Julius: resolve the badly-soldered JLC USB connectors.
  • Julius: post a tethering drawing for review.
  • KBM: post safety and reflective sticker suggestions.