← Back to the journal

Build 001 · Freestyle · Grounded · video-path diagnosis

Source One V5 · 4S

A real 4S analog freestyle build based on the TBS Source One V5, now grounded while an intermittent frozen FPV image is diagnosed across the camera and analog video path.

DB–001Repository record
Earlier bench photograph of the TBS Source One V5 with a blue printed GoPro mount above the analog FPV cameraEarlier bench configuration · GoPro and mount have since been upgraded
Airframe
5-inch
Battery
4S · 1500mAh
Video
Analog · 5.8GHz
Dry weight
Unknown

A repairable 5-inch freestyle quad built as a hands-on platform for learning assembly, soldering, and FPV flight.

This build pairs a TBS Source One V5 frame with a 4S analog power and video system. It is the first real aircraft recorded in Drone Builder: not just a parts list, but a record of the knowledge and practice required to turn those parts into something flyable.

ChatGPT was used extensively during the build. Bacon Ninja FPV’s Source One material was especially useful for assembly and soldering even though the reference build was not identical; James remembers the video transmitter as the main variation. Joshua Bardwell’s build instruction and flight-learning videos were also important. Learning to fly took sustained practice rather than ending when assembly was complete.

Parts used together.

RoleComponentEvidence
FrameTBS Source One V5verified
MotorsEMAX ECO II 2306 1900KV × 4verified
Flight controller + ESCSpeedyBee F405 V3 BLS 60A 30×30 FC & ESC Stackverified
Control receiverHappymodel EP1 2.4GHz ExpressLRS nano receiververified
Video transmitterRushFPV Rush Tank SOLO 1W 5.8GHz VTXverified
FPV cameraRunCam Phoenix 2 Special Editionverified
Video antennaRushFPV Cherry2 5.8GHz RHCP antenna · straight MMCXverified
PropellersHQProp Ethix S5 Light Grey 5×4×3 tri-bladeverified
BatteryCNHL Black Series 1500mAh 4S 14.8V 130C XT60verified

Used with the aircraft.

Action camera · verified

GoPro HERO13

Current action camera. The historical photographs show the earlier GoPro and high-vibration blue printed mount.

Details worth finding again.

Receiver identityRecorded as a Happymodel EP1. The original note also contained PP/SX1280 wording, so the label on the installed receiver is worth checking.
Capacitor placementMounted in the limited space between the FPV camera and flight controller, following the layout used in Bacon Ninja FPV’s tutorial.
Motor-wire geometryThe motor wires were soldered from the outside of the ESC pad area. In hindsight, approaching from the inside would have produced better geometry; the full lesson still needs to be written.
Early GoPro mountThe blue printed camera and mount configuration visible in the historical photos produced substantial vibration and was replaced.
Printed parts archiveThe GoPro mount STL is retained. Bumper, antenna mount, motor guards, ELRS holder, and battery holder models can be added with their sources, licenses, materials, slicer settings, and revisions.
ConfigurationBetaflight version, receiver setup, rates, ESC firmware, and tune are not yet recorded.
Flight learningLearning to fly took a long time. Joshua Bardwell’s instructional videos were especially helpful.
Learning footageFootage from the flight-learning phase exists and can be attached later. No footage of the original assembly is currently available.

Failures recorded without rewriting uncertainty.

Raw recordings remain outside Git. This record keeps the source identity, checksum, event timing, direct observations, pilot recollection, and current interpretation together.

VTX-LOSS-0022025-09-27
open

Safety-critical · loss of pilot visual control

In-flight frozen FPV image followed by uncontrolled descent

The pilot observed the FPV image freeze at approximately 0:06. The GoPro records controlled flight immediately beforehand, followed by a sustained tumble and ground impact. The image did not recover until the aircraft was rebooted.

Evidence timeline

  1. GoPro shows controlled flight with no obstacle contact visible.
  2. Aircraft abruptly begins a sustained tumble, closely aligning with the reported FPV freeze.
  3. Ground impact; aircraft comes to rest.

Direct observations

  • The GoPro continues recording normally throughout the event.
  • No collision is visible immediately before the aircraft departs controlled flight.
  • No goggles DVR has yet been found for this incident.

Pilot report

  • FPV image froze at approximately 0:06.
  • The symptom matched the later June incident.
  • The live image did not recover until the aircraft was rebooted.

Current interpretation

  • This is the stronger evidence that an immediate impact is not required to trigger the failure.
  • The GoPro establishes the aircraft behavior and timing, but cannot independently prove what remained active in the goggles feed.
  • Exact failed component remains unresolved.
Retained source files & checksums
GoPro context clipvtx-loss-002-gopro-context-trim.mp4MP4 · H.264 · 3840×2160 · 29.97 fps · 14.400 seconds

Preserves controlled flight, onset of the sustained tumble, and impact. The embedded creation date reflects the later trim/export, not the flight date.

SHA-256 2beba03ea7cdae264a83a61655c2b84815195d994744f70139899b573de9366a
VTX-LOSS-0012026-06-11
open

Diagnostic evidence · OSD and RF remained active

Impact-associated camera or video-path freeze

Synchronized GoPro and goggles recordings show the camera background freeze at approximately the time of impact while Betaflight OSD continued to update. RF transmission persisted until aircraft power was interrupted, and live video returned only after reboot.

Evidence timeline

  1. Major roll and impact. Audio alignment places this close to the goggles freeze.
  2. Camera background stops updating abruptly.
  3. Betaflight post-flight statistics update over the frozen camera background.
  4. RF static appears when aircraft power is interrupted.
  5. Betaflight boot screen appears, followed by restored live video.

Direct observations

  • The camera image froze, but live OSD changes remained visible over it.
  • The goggles continued receiving the aircraft transmission until power interruption.
  • Post-flight values appear to show approximately 14.92 V minimum battery, 12 A maximum current, 40 mAh used, and 00:19 arm time.
  • The available evidence does not show an unusually large throttle or current event at the failure.

Pilot report

  • An earlier crash occurred in the longer raw footage.
  • The frozen image did not recover until the aircraft was power-cycled.

Current interpretation

  • This was not a complete VTX or RF dropout: transmission and flight-controller OSD remained active.
  • The likely fault boundary includes the FPV camera, its power/ground/video wiring, or the flight-controller video-input path.
  • Because VTX-LOSS-002 occurred without an immediately preceding impact, impact may trigger the fault but is not required.

Configuration observed in evidence

  • SmartAudio menu initially showed Raceband 8 at 5917 MHz.
  • Displayed VTX power was 800 mW before later menu inspection changed the selected values.
  • PIT mode displayed OFF.
Retained source files & checksums
Goggles DVR contextvtx-loss-001-goggles-context.aviAVI · MJPEG · 640×480 · 29.9 fps · 271.098 seconds

Preserves the frozen camera background, continuing OSD, SmartAudio inspection, loss of RF at power interruption, and restored live video after boot.

SHA-256 98d13558d6a43d0c411df872066695416a88a8c8161d3650ba2cf9e658fd5562
GoPro context clipvtx-loss-001-gopro-context-trim.mp4MP4 · H.264 · 3840×2160 · 29.97 fps · 61.547 seconds

Preserves aircraft motion and impact for approximate alignment with the goggles DVR. The embedded creation date reflects the later trim/export.

SHA-256 ca78008dda8a54d8805f0a3e383a3c146782ad7a3b968404e0c8df99d36a826b

The knowledge behind the build.

used

Soldering

Joining and inspecting the power, motor, receiver, camera, and video-transmitter connections required by a custom build.

used

Frame & electronics assembly

Turning separate airframe and electronics components into a serviceable working quad.

learning

FPV flight

A substantial learning phase after the build itself; tutorial guidance and repeated practice were both important.

Simulator training

The physical build has source files too.

retained

Source One V5 GoPro mount

Binary STL · 9,292 facets · 454 KB

Supplied with the build record for the HERO13 mount. Source attribution, license, print material, slicer settings, and installed revision should be recorded before it becomes a public download.

artifacts/tbs-source-one-v5-4s/source-one-v5-gopro-mount.stl

State changes, not overwritten history.

  1. Date unknown

    Initial build

    Assembled the Source One V5 around the SpeedyBee stack, EMAX motors, analog video system, and ExpressLRS control link.

  2. 2025-07-15

    Stack and motor-wire assembly

    Progress photos preserve the loose-lead state and the later same-day routing and soldering around the outside of the ESC pads. The geometry became a lesson worth revisiting.

  3. 2025-07-22

    Early GoPro configuration

    The assembled aircraft was photographed with props, battery strap, and the blue printed GoPro mount that later proved to have substantial vibration.

  4. 2025-08-03

    Earlier bench configuration

    Photographed on the bench with the earlier GoPro and blue printed mount. The capacitor was packaged between the FPV camera and flight controller because space was limited.

  5. Date unknown

    GoPro & mount upgrade

    Replaced the earlier high-vibration action-camera setup with a GoPro HERO13 and a different printed mount. The model file is retained; print settings and installed revision remain to be recorded.

  6. After initial build

    Flight-learning phase

    The aircraft became a continuing practice platform. Most simulator training happened in DRL; Liftoff is now used for continued practice.

  7. 2025-09-27

    VTX-LOSS-002 · in-flight image freeze

    The FPV image reportedly froze during otherwise controlled flight. GoPro evidence shows the aircraft enter a sustained tumble at approximately the reported freeze time and impact several seconds later. No immediately preceding collision is visible; recovery required an aircraft reboot.

  8. 2026-06-11

    VTX-LOSS-001 · OSD continued over frozen image

    Companion GoPro and goggles recordings captured another nonrecovering freeze. Betaflight OSD and RF transmission remained active over the frozen camera background, narrowing the diagnostic boundary away from a complete VTX dropout.

  9. 2026-08-24

    Incident evidence correlated

    Preserved clip metadata, checksums, synchronized event timing, observations, pilot reports, and diagnostic interpretations for both known frozen-image incidents. Raw video remains outside Git.

  10. Current

    Grounded for video-path diagnosis

    At least two nonrecovering FPV image freezes are now documented. Because dependable live video is required for safe flight, the aircraft remains grounded until the camera/video-path fault is isolated, corrected, and verified.

  11. Current

    Bench modification pass

    The aircraft is currently on the bench undergoing modifications. The parts changed, reasoning, and result will be added as the work progresses.