Air65 — Bench Setup and Preference Port

Setup

A BetaFPV Air65 joined the workshop. The goal was to make it fly like the existing craft without importing anything specific to that airframe.

Read over USB with the board on the bench, no battery and no props. The factory configuration was backed up before any change.

What the factory config had wrong

Area	Factory	Target
ARM	AUX1, 1700–2100	AUX4, 1800–2100
BEEPER	AUX4	AUX1
Crashflip	slot 4, AUX3	slot 5, AUX5
Blackbox mode	absent	AUX1, full span
OSD profile adjustment	absent	AUX3
Rate system	`BETAFLIGHT`, 110/74/0.30	`ACTUAL`, 170/550/0.35
Throttle	30/0/34, no limit	40/35/44, 80% scale
Telemetry	off	on

Arm and beeper were on each other’s switches. Everything else on this list is comfort; this one is a safety problem. A craft that arms on the switch your muscle memory uses for the beeper will arm when you did not ask it to.

What changed

Sixty-seven commands in a single saved session, after a backup. Ported:

The full switch map, slot for slot, including the cleared slot 4.
Actual rates — the same 170/550/0.35 and 120/400/0.25.
The throttle curve and its 80% scale limit.
The OSD layout, all 32 element positions plus both bitmasks.
Crashflip rate and auto-rearm, and the VTX power switch table.
fpv_mix_degrees = 10, pilot_name, and the telemetry feature.

Deliberately not ported: PIDs, filters, motor_idle, align_board_yaw, acc_calibration, and the VTX frequency. Those describe the airframe, not the pilot.

Findings

The aux, adjrange and rateprofile 0 blocks now come back byte-identical to the other craft’s backup.

rates_type is absent from diff afterwards, which is the confirmation that matters. A diff prints only non-default values, so its absence proves the setting equals the firmware default — the same place the other craft sits. A craft still showing set rates_type = BETAFLIGHT has not been converted.

Pre-flight audit passed on bi-directional DShot, receiver protocol and blackbox flash. The one warning, RXLOSS DSHOT_TELEM, is expected with the radio off and no battery.

Open issues

Copying the OSD layout initially carried vcd_video_system = AUTO across with it. That is a board setting, not a pilot setting, and AUTO is the state that blanks the screen on these boards. Reverted to NTSC in the same session and verified by reading it back. See Gyro Substitutions and Firmware.

Still to do before this craft is trusted:

Confirm with the radio on and props off that ARM reads above 1800 on AUX4.
Check the OSD in the goggles. The element rows came from a different craft.
First hover and a blackbox log, for a motor balance and gyro noise baseline.
Revisit motor_idle = 550, against 400 on the other craft.

Method

No blackbox log for this entry — it was a bench session. Configuration was read and written over USB serial with the CLI, which remains the reliable transport on this firmware.