Notas de la versión

1.4.1-beta.37

FFB-Bridge 1.4.1-beta.37 corrects autopilot following on official MOZA flight bases and includes the recent X-Plane telemetry fixes.

• MOZA autopilot follow now follows the aircraft's real control surfaces. On MSFS, beta.34 could leave an AB6, AB9, or AY210 physically centered while the autopilot banked or pitched the aircraft. The bridge was reading pilot input instead of AP-driven surface deflection. Official MOZA devices now use the aircraft's actual elevator and aileron deflection, with the flight-director reference retained only as a fallback when surface telemetry is unavailable.
• X-Plane airspeed is correct again. Approximately 300 kt could appear as roughly 2078 kt because X-Plane 12 supplied Qstatic in Pascals while its official catalog labels the value as psf. The bridge now handles the live unit correctly before the value reaches the dashboard or force calculations.
• Retracted spoilers no longer produce spoiler activity or yoke rumble. The bridge now reads X-Plane's official actual-speedbrake surface ratio. A reported zero is authoritative and no longer falls back to the cockpit handle position.
• All production X-Plane datarefs have been audited. Automated catalog checks now prevent unknown or malformed datarefs from entering future builds. The audit also corrected rotor telemetry, retained RPM-derived engine rumble where X-Plane has no native vibration dataref, and preserved X-Plane 11 flap compatibility.
• Safety and existing control behavior are unchanged. Force trim, input-isolated MOZA follow conditioning, heartbeat expiry, STOPALL cleanup, and legacy-device autopilot follow retain their existing behavior.

1.4.1-beta.36

FFB-Bridge 1.4.1-beta.36 fixes the incorrect X-Plane airspeed introduced by the dynamic-pressure change in beta.34.

• 300 kt no longer appears as roughly 2078 kt. X-Plane 12's live Qstatic value is in Pascals even though Laminar's official dataref catalog labels it as psf. The bridge converted that value from psf to Pascals a second time, inflating airspeed by about 6.92×.
• Force feel is corrected too. The bad value was in the shared flight snapshot, so it could make aerodynamic loading and airspeed-gated drag, buffet, and vibration effects reach their high-speed ranges far too early. The adapter now normalizes X-Plane's runtime unit before any dashboard or force calculation sees it.
• The boundary is resilient to simulator-version differences. The bridge defaults to the field-proven Pascal convention and can recognize and latch either Pascal or documented-psf behavior while indicated airspeed is healthy. A frozen or covered pitot cannot later change that choice.

1.4.1-beta.35

FFB-Bridge 1.4.1-beta.35 corrects X-Plane spoiler activity and audits every X-Plane data source used by the bridge against Laminar Research's official catalog.

• No spoiler rumble while the surfaces are retracted. The bridge was asking X-Plane for two inferred left/right spoiler datarefs that do not exist. Their zero responses were treated as unavailable, causing the cockpit speedbrake handle to be substituted for the physical surfaces. This could show spoiler activity and shake the yoke even though the actual spoilers were at zero. The bridge now reads X-Plane's official aggregate actual-surface ratio, and a reported zero is authoritative.
• The entire X-Plane dataref list has been verified. Production, diagnostics, the experimental mapper, and datarefs written for axis ownership were checked for exact name, type, units, array dimensions, read/write access, and supported X-Plane versions. Automated checks now block unknown or malformed datarefs from entering future builds.
• Rotor telemetry is corrected. Main-rotor speed now comes from actual flight-model rotation rather than the prop-handle request, and the native blade-slap source is correctly read as a per-rotor array.
• Engine rumble remains available on X-Plane. A nonexistent stock engine-vibration dataref has been removed; X-Plane continues using the existing RPM-derived rumble fallback.
• X-Plane 11 flap handling is preserved. The bridge explicitly uses the official legacy flap-handle source on XP11 while preferring the current source on XP12.

1.4.1-beta.34

FFB-Bridge 1.4.1-beta.34 changes where force feel gets its airspeed on both simulators. Forces now follow the aircraft's actual aerodynamic state instead of the airspeed indicator.

• Pitot failures no longer affect force feel. Previously the aero forces (control loading, stall buffet, overspeed, glider feel) were driven by the simulator's indicated airspeed, which reads through the simulated pitot-static system. A field report on X-Plane showed the result: starting a flight with the pitot cover on left the yoke with no aero feel at all, and reapplying the cover at speed froze the indicated airspeed so the yoke kept full high-speed resistance while parked on the ground. Both simulators now feed the force model from the flight model's dynamic pressure, which a covered or failed pitot cannot change. The instrument can fail; the feel stays true to the airflow.
• Normal flying feels the same. At typical general-aviation speeds and altitudes the derived value matches the indicated airspeed to about a knot. At high altitude and high speed the forces now track the true aerodynamic load, which sits a few knots below the indicated value there.
• If the simulator does not supply dynamic pressure, the bridge falls back to the previous indicated-airspeed behaviour automatically.

1.4.1-beta.33

FFB-Bridge 1.4.1-beta.33 is a major update for MOZA flight bases: trim and autopilot following are now carried out by the base itself, the autopilot Authority slider opens to the full range on both simulators, and helicopter force trim anchors the cyclic exactly where it is released. It also includes the Logitech G940 autocenter corrections from the beta.14 to beta.16 diagnostic builds.

MOZA flight bases (AB9, AY210, AB6):

• The autopilot can now really fly your controls. With autopilot follow enabled, the yoke or stick physically follows the aircraft's actual control movements: forward and holding forward through a descent, banking with a turn, and easing back only as the autopilot eases the controls. The Tuning → Autopilot → Authority slider now goes to 100% on both simulators and is the only control over how far the follow moves.
• MSFS no longer fights the follow. Follow motion on MOZA bases is no longer visible to MSFS as pilot input, so high authority no longer risks disconnecting the autopilot, and the previous 10% live cap on MSFS is gone on these bases. Your own hand inputs still reach the simulator normally, including pushing through the autopilot.
• Trim physically moves the controls. Trimming walks the yoke to the new trimmed position and holds it there. The stick settles at trim rather than drifting near it.
• Helicopter force trim anchors exactly. With the trim release button held the cyclic is completely loose; on release it becomes firm at that exact spot, with no drift or creep, and it holds firmly at the trimmed position rather than only resisting displacement. Beep trim physically steps the cyclic to each new position.
• New "Follow speed" slider (Tuning → Autopilot) sets how quickly the base walks the controls to trim and autopilot positions. 100% is the fastest and is the default.
• Fail-safe behaviour: if FFB-Bridge stops for any reason, the base clears the follow positions on its own within about three seconds.

X-Plane: autopilot follow now tracks the aircraft's control-surface deflections on MOZA bases, giving the same held-through-the-descent behaviour as MSFS. Aircraft that do not report surface deflections keep the previous flight-director following.

Logitech G940: corrected the DirectInput autocenter handling. The autocenter setting is now applied at the documented point in the device lifecycle, read back and verified, and reported accurately in support bundles.

All other devices are unchanged, including the existing autopilot-follow behaviour and limits on non-MOZA sticks.

1.4.1-beta.13

FFB-Bridge 1.4.1-beta.13 makes the autopilot hold the controls the way it was always meant to, after a user report that the yoke went soft with a wide free travel the moment the autopilot engaged.

• The autopilot now holds the yoke firmly. The autopilot-follow hold was rendering as a gentle linear spring, so close to the autopilot's position it offered almost no resistance. It now uses the same magnetic-brake style hold as helicopter force trim: firm resistance begins immediately when you displace the controls from where the autopilot has them, and on MOZA bases the hold uses the device's full strength rather than being limited by the flight-force ceiling. The Tuning → Autopilot → Strength slider sets how firm the hold is and how much force it takes to override it; Master gain still applies.
• The same firmer hold applies to X-Plane's autopilot following, and the hold no longer softens momentarily while FFB-Bridge is backing off to avoid fighting the MSFS autopilot.

How far the yoke travels with the autopilot is unchanged (the Authority slider, still capped in live MSFS flight so the autopilot does not disconnect).

1.4.1-beta.12

FFB-Bridge 1.4.1-beta.12 adds the trim-reset function helicopter pilots know from SimFFB's "Center" button, requested by a user migrating from it, and fixes a hidden helicopter setting on MOZA bases.

• A new "Trim reset button" assignment (Hardware → Helicopter, next to the FTR and force trim on/off buttons). With force trim engaged away from centre, one press cancels the trimmed position: the hold pulls the cyclic back to the centre of the axes and stays engaged. Trim again with the FTR button as usual. Like the other force-trim controls, it can be learned on any connected USB controller, not only the force-feedback device.
• The "Force trim on/off button" row is visible again on MOZA bases. Since 1.3.2 it was accidentally hidden on machines with a MOZA base, so the pedestal-switch stand-in could not be assigned from the UI there. Existing assignments kept working; only the setup row was missing.

1.4.1-beta.10

FFB-Bridge 1.4.1-beta.10 fixes force feedback on Linux for the Logitech Force 3D Pro, WingMan Force 3D, and Flight System G940, and improves support-bundle accuracy.

• Constant-only Logitech sticks now arm on Linux. The Linux kernel drivers for these three sticks render constant force only and reject the centering-spring effect the bridge uploaded at arm time, so arming failed in a repeating loop and no forces were ever produced. The bridge now reads the driver's real capability list, skips the unsupported spring, and renders centering as a constant force computed from the stick position instead. Periodic effects (rumbles and buffets) are not supported by these drivers in hardware mode; enable Hardware → Effects → "Effect rendering: Software" to get them synthesized.
• A repeated arming failure can no longer fill the device's effect table. On Linux, each failed arming attempt used to leave its partially created effects behind; after enough retries the device ran out of effect slots until the app was restarted. Retries now free those effects first. Reported with a full kernel-level analysis in a Force 3D Pro report — thank you.
• Support bundles no longer name a default device when none was opened. A bundle exported while no force-feedback stick was open used to claim device 045E:001B (SideWinder FFB2) in its README, and its Windows device listing filtered on that same wrong vendor and captured nothing. The README now says no device was opened, and the device listings widen to every supported force-feedback vendor so the bundle shows the hardware that failed to open.

Force feel on Windows, macOS, and spring-capable Linux devices is unchanged.

1.4.1-beta.9

FFB-Bridge 1.4.1-beta.9 improves support diagnostics, prompted by a beta report that took four support bundles to trace to a profile-selection difference.

• The session log now records which tuning profile is applied and why. Every profile change is logged, and automatic selection also records what it matched on — your saved binding, a saved profile's name, a built-in starter, or the default fallback.
• Support bundles now include your aircraft-to-profile assignments and a list of your saved profiles (file names, sizes, and titles — not their tuning values). With these, a single bundle can answer "which tuning was active for this aircraft, and where did it come from".

Nothing about force feel changes in this build.

1.4.1-beta.8

FFB-Bridge 1.4.1-beta.8 makes several effects read the aircraft's real numbers instead of fixed ones, and adds a way to test forces across the speed range without starting a flight. (Beta.4 through beta.7 were internal candidates; this build rolls them up.)

• When something else is holding the SimConnect port, FFB-Bridge now says what it is. If connecting to MSFS repeatedly gets an answer that is not SimConnect — most often a Flight Simulator process left over from a previous session that never fully exited — the log now names that process and its PID instead of retrying silently, with the suggestion to end it in Task Manager. Diagnosed from a user report where only a reboot appeared to fix repeated "simulator not detected" sessions.

• The Support page's "MSFS reachable" check now verifies SimConnect is actually answering, instead of reporting a pass whenever anything accepts the connection. A port held by a non-simulator process now shows a warning naming that process rather than "MSFS is running".

• A new "Engine idle roughness" control (Tuning → Effects) adds extra shake at idle RPM that fades out as power comes up, on top of the engine rumble. The Cessna 172 starter now uses a small amount of it by default; every other profile starts at 0.

• Rotor rumble now follows the rotor, not the engine switch. Shutting the engine down used to silence the rotor rumble instantly even though the rotor was still turning. The rumble — and the forward-flight per-rev vibration — now keep rendering while the rotor turns and fade as it winds down, so shutdown spool-down and autorotation feel right. Reported by a light-helicopter user.

• The retreating-blade-stall buffet now uses your helicopter's own VNE. It previously used a fixed speed carried by the profile, so a fast helicopter flown on a starter built for a slower one started shaking well before its actual redline — reported on an S-76 shaking from about 110 knots. FFB-Bridge now reads the redline the simulator reports for the aircraft you are flying, and falls back to the profile value when the simulator does not publish one. Helicopters without their own starter profile benefit most.

• A new "Blade-stall onset" control (Tuning → Advanced → Helicopter effects) sets how close to VNE the buffet begins. It was previously editable only by hand-editing profile files.

• Helicopter vibration now matches your rotor. The signature forward- flight shake is one beat per blade per rotor revolution, so a two-blade Bell 206 thumps slowly while a three- or four-blade machine buzzes faster and lighter. FFB-Bridge now derives that from the aircraft's blade count and live rotor speed, so spool-up and rotor droop are felt too. The Bell 206 is unchanged by design — it is the aircraft this effect was tuned on. Helicopters using the generic trim profiles now assume a three-blade rotor, which suits far more of the fleet than the previous two.

• Engine rumble can follow engine speed. On the Cessna 172 starter the rumble is now rougher at idle and smooths out in the cruise, and its frequency rises and falls with RPM instead of staying fixed — closer to what a real pilot feels through a yoke. Other profiles are unchanged for now; the option is per-profile.

• Flight Check can now sweep airspeed and RPM. Steps that depend on speed — airspeed load, engine rumble, ETL, two-per-rev, retreating blade stall, rotor spool-up — show a slider while running, so you can feel how a force builds across its range without loading a flight. Leaving the slider alone reproduces the previous fixed test exactly.

• MSFS installed on a second drive is now found automatically. If your Steam library is not on the default drive, FFB-Bridge reads Steam's own library list instead of giving up. If your install is somewhere it still cannot find, Support → Health checks now offers Locate manually… to point it at the SimConnect.xml or the Steam library folder.

Also in this build, from work already on the development line: force-trim, trim-toggle and beep-trim controls can be bound to any USB controller, not only the force-feedback device — for grips and button boxes fitted to modified sticks.

Helicopter pilots: the vibration and blade-stall changes want real flying before 1.4.1 goes stable. Please report through ffb-bridge.com/feedback or the community.

1.4.1-beta.3

FFB-Bridge 1.4.1-beta.3 answers helicopter force-trim reports from MOZA AB6 users, and stops the automatic Safety Stop firing on sticks it cannot protect.

• Force trim now holds with the full strength your device can produce. The hold was being limited by the physical output ceiling meant for active flight forces — on an AB6/AB9 that capped it at roughly a third of the base's capability, which is why it felt like a rubber band and why moving the Force trim hold slider barely changed anything. A trim hold only resists you moving the cyclic, so it now commands its real strength, like the device's damping and friction already did. Master gain still applies.
• The cyclic is stable across its full travel again. In 1.4.1-beta.2 the aerodynamic load could grow as you approached the travel limits and push the stick further out rather than resisting it. Force trim now measures your displacement in the simulator's own control axis rather than mixing it with the device's physical position, and only once the trim brake has actually engaged. This one is still waiting on AB6 owners to confirm in flight — if you fly helicopters, please report what you feel.
• The automatic Safety Stop is now matched to your device. It stays on for strong bases such as MOZA and Brunner, and for any device FFB-Bridge has not characterised. It is off by default for the legacy consumer sticks — SideWinder Force Feedback 2 and Force Feedback Pro, and the supported Logitech sticks — whose motors cannot produce the runaway forces it exists to interrupt, and where it mostly produced nuisance stops during brisk control inputs.
• You can now turn the Safety Stop on or off yourself, under Settings → Session. Turning it off asks you to confirm first and explains what you give up; your choice is remembered even if you change device. Observation continues either way, so force trim and cyclic damping are unaffected, and a stop that has already fired still needs an explicit re-arm.
• Settings is now organised into tabs — General, Session, Updates and Local data — instead of one long page.

This build includes all 1.4.1-beta.2 and beta.1 corrections below. Helicopter pilots on MOZA hardware are the priority testers for this one: please report through ffb-bridge.com/feedback or the community.

1.4.1-beta.2

FFB-Bridge 1.4.1-beta.2 makes the automatic Safety Stop intervene much sooner when a powerful yoke or stick enters a sustained runaway oscillation.

• A large, non-decaying oscillation now cuts force after four cycles instead of eight. The detector requires 8 confirmed turns rather than 16, halving the dangerous motion it must observe before setting device gain to zero and issuing STOPALL.
• The false-positive protections remain intact. The near-full-travel amplitude requirement, turn hysteresis, three-second evidence window, decaying-ring rejection, latched shutdown, and explicit re-arm requirement are unchanged. Normal control movement, hand tremor, small vibration, three brisk large cycles, and a settling spring release do not qualify.
• The slower heavy-yoke case is now covered. At the production detector's 200 Hz cadence, deterministic tests cut sustained ±0.8 travel at 1.105 seconds at 4 Hz and 2.205 seconds at 2 Hz. The former 16-turn requirement could not fit a 2 Hz runaway inside its three-second window.

This build includes all 1.4.1-beta.1 corrections below. AY210 and other high-torque yoke users: please test at your normal roll-force setting and report both the stop response and any unexpected nuisance trip through ffb-bridge.com/feedback or the community.

1.4.1-beta.1

FFB-Bridge 1.4.1-beta.1 is a focused public test build for two field-reported control problems, plus the completed canonical-direction correction.

• X-Plane: FFB-Bridge no longer resets another plugin's joystick override datarefs while it is not using axis ownership. Pitch and roll override are now written only while FFB-Bridge's own Autopilot Follow pass-through is active. When that owned session ends, FFB-Bridge releases both flags once; an inactive bridge leaves them untouched. This specifically needs testing with aircraft that manage their own overrides, including the UH-60M.
• Helicopter force trim now establishes one coherent equilibrium. The captured cyclic datum is used by both the hold spring and the rotorcraft aerodynamic load, so the aerodynamic term no longer pulls the control back toward geometric centre after FTR release. A release without a valid current axis sample stays loose instead of engaging a stale/default centre. The existing per-device output safety ceiling remains in force. This correction needs AB6 testing with the Bell 206 and other force-trim helicopters: move the cyclic, release FTR, and report whether it remains at the full released position without a pitch-up pull.
• All directional force output now uses one canonical convention and one hardware boundary. Source vectors remain right-positive/aft-positive through composition, and device compatibility is applied once at the final encoder. The normal Windows raw HID/PID routes passed physical cardinal, spring, trim, damping, and extended checks on MOZA AB9, MOZA AY210, and SideWinder FFB2; AB9 DirectInput passed separately. Saved legacy polarity settings are migrated without adding a second downstream sign correction.

Install this beta through the beta update channel or download it from ffb-bridge.com. Please report the aircraft, simulator, device/backend, and whether the specific result passed at [email protected] or in the community.

1.4.0

FFB-Bridge 1.4.0 launches the aircraft library at ffb-bridge.com and rebuilds every built-in profile from documented sources. The goal of both is the same: get the community involved in making these profiles genuinely good.

• The aircraft library is live at ffb-bridge.com/aircraft. This is the centre of this release. Every built-in aircraft now has a public page showing what the profile models, the sources behind each claim, its tuning values, and where the gaps are. You can share your own tuned profiles straight from the app (the globe button in Profiles), and a shared profile shows exactly what it changes relative to the baseline it started from. That difference is the valuable part: when someone who knows an aircraft adjusts a profile, everyone can see what they changed and why it's better. Corrections, shared profiles, and flight reports all feed the next revision of the built-in profiles.
• All 13 built-in profiles were rebuilt from documented evidence, and they need your experience. Each profile is now grounded in POH and certification data, flight-test reports, and pilot accounts: the C172 is trainer-heavy in pitch, the Extra 300 stays light from stall to VNE, the P-51D carries a strong speed load with a light per-g gradient, the transports have much firmer yokes. But documents only go so far. These profiles are built from sources, not type experience. If you have real-world time in one of these aircraft and the profile doesn't match what your hands remember, say so on that aircraft's library page or share a corrected profile. That input is exactly what the library exists for.
• X-Plane: Autopilot Follow can now be set up to 50%. While the autopilot flies, FFB-Bridge uses X-Plane's per-axis override so the simulator reads your hand input from the bridge instead of the raw joystick axis. This lets the yoke follow the autopilot with more authority without the simulator treating that motion as pilot input, and moving the yoke yourself still reaches the aircraft. This feature is new in this release and has had one day of bench testing on a single MOZA AY210: the motion is usable but not yet fully smooth, and brief counter-movements of the on-screen yoke can still appear during fast corrections. Start around 15–25%, increase to taste, and report what you find.
• MSFS: Autopilot Follow stays capped at 10%. MSFS reads the physical axis directly, so above roughly that level the simulator's autopilot and the follow cue can fight each other. Live testing showed the automatic back-off added during this cycle softens the fight but cannot stop a divergence at high authority. Saved profiles above 10% are clamped in live flight.
• MOZA AB6 with helicopters: a fix for the periodic jerk and rough centre. Two field reports traced to FFB-Bridge's background device-status traffic briefly interrupting the AB6's own damper while a rotorcraft profile was active. That traffic is now reduced to the safety-essential minimum while forces are live, and diagnostic position/torque polling runs only while the Hardware page is open. This fix could not be verified on an AB6 in-house, so reports from AB6 helicopter pilots are needed, positive or negative.
• MOZA AY210: no more unexpected reference motions, and a per-axis output envelope. The base no longer replays its startup routine during Flight Check or re-arm, Disarm is an absolute force-off on every path, and the automatic envelope is now 35% pitch and 50% roll to match the yoke's different pitch and roll mechanisms. Explicitly saved limits are preserved.
• PicoWinder is now supported out of the box on Windows and Linux, with the full SideWinder-class output range.
• A crash in the simulator now produces one bounded impact through the existing touchdown channel, within your configured force limits, and it does not repeat. A frozen or disconnected simulator behaves as before: forces off, no synthetic impact.
• The 737 stick shaker works on more add-ons. Aircraft that do not publish the standard stall-warning value get the shaker from an airborne-only angle-of-attack fallback, and the 737's shaker and buffet levels were raised to match their certified character.
• The AS 33 sailplane starter now ships with vertical-wind vibration enabled (progressive over a 1.5–15 ft/s envelope), so lift and sink are feelable. The channel stays off by default in other profiles.
• Assorted fixes: enabling Autopilot Follow can no longer soften a heavier aircraft profile's centering; engine telemetry (RPM, vibration, reverse) covers every installed engine instead of engine 1; explicit flap motion and arrival cues across the fleet.

Update from within the app, or download from ffb-bridge.com. Feedback on the rebuilt profiles, the X-Plane autopilot follow, and the AB6 fix: [email protected] or the community.

1.4.0-beta.4

FFB-Bridge 1.4.0-beta.4 sets the Autopilot Follow ceiling back to 10% after live validation of the 50% range.

• Autopilot Follow now goes up to 10%, not 50%. Live testing showed what the higher range actually does: the simulator reads the motor-driven yoke as pilot input, its autopilot counters, and above roughly 10% the two can over-correct each other until the aircraft is upset — the automatic back-off added in beta.2 softens this but cannot stop a divergence once it starts. 10% is the strongest setting that behaves in real flights. Saved profiles with higher values are clamped automatically.
• The beta.2 protection stays, quieter. The automatic back-off now reacts only to a genuine sustained fight, not to the autopilot's ordinary small corrections, so "AP Follow reduced…" log lines should be rare.
• The Flight Check autopilot spotlight is unchanged — it runs on the bench with no simulator in the loop, where the stronger demonstration cue is safe.

1.4.0-beta.2

FFB-Bridge 1.4.0-beta.2 makes high Autopilot Follow settings degrade gracefully instead of fighting the simulator's autopilot. Field report: at 50% authority the yoke and the in-sim yoke could end up moving in opposite directions until the aircraft departed — the simulator reads the bridge-driven axis as pilot input, and its autopilot counters what it thinks you commanded.

• Autopilot Follow now moves the yoke smoothly. The followed position is rate-limited and smoothed, so the cue can no longer step-drive the axis (which both feels wrong and is exactly what the simulator's pilot-override detection reacts to).
• A new opposition governor watches for the autopilot fighting the cue. When the actual control surfaces persistently counter the followed position, the bridge automatically backs Follow authority down toward the safe 5% cue, and logs why (see the session log: "AP Follow reduced…"). Authority recovers on its own once the opposition clears. Settings at or below 5% are never reduced.
• What this means in practice: you can now explore higher Follow values safely — if your aircraft's autopilot pushes back, the bridge yields instead of escalating into a spiral. Full 1:1 yoke mirroring under autopilot remains impossible while the simulator can read the physical axis; the governor finds each aircraft's usable ceiling automatically.

1.4.0-beta.1

FFB-Bridge 1.4.0-beta.1 is the first public beta of the profile-library release. It rolls up the unpublished 1.3.2-beta.1 through beta.7 local candidates (the AY210 lifecycle correction, the bounded roll-authority and Autopilot Follow work, the aircraft-impact cue, and PicoWinder first-class support) and adds the research-grounded retune of every built-in aircraft profile.

• All 13 built-in profiles retuned from cited control-feel research. Every starter magnitude now traces to a published source — pilot flight tests, certification data, and primary-source measurements (including wartime NACA flight-test data for the P-51D and official Airbus sidestick force tables for the A320). Direct-cable aircraft now use a physical V² airspeed-load law, the centering force keeps firming with speed instead of going flat above rotation, and each aircraft carries its documented pitch/roll harmony.
• Notable per-aircraft corrections: the Extra 300 is now light and nearly speed-invariant (its aerodynamic balancing is the documented character — the old strong q-ramp was wrong); the P-51D is light per-g with forces that build strongly with speed; the C172 carries the firmest per-g yoke of the trainers, as the real one does; the DA40's pronounced stall buffet and lighter overall forces are modeled; the King Air 350's stall warning is now correctly the horn (the production aircraft has no stick shaker); the TBM 930's shaker is stronger to match its documented "vigorous" cue.
• Aircraft library. Each release now ships per-aircraft dossiers — the research, the force mix, and the canonical starter baseline — which the ffb-bridge.com aircraft pages render. Profiles you save now record which starter they were forked from, so shared profiles can show exactly what you changed and why.
• How the profiles were built is public. The methodology, sources, and per-value citations for every starter are in the aircraft library on the website.

Feedback on the new profile feel is the point of this beta — the Feedback page (or a profile shared to the library with your changes) is the fastest way to reach us.

1.3.2-beta.5

FFB-Bridge 1.3.2-beta.5 is the cumulative local hotfix candidate for the AY210 lifecycle correction and the 1.3.2 profile/effect follow-up. It remains unsigned and unpublished.

• PicoWinder is now a first-class supported device. The open-source PicoWinder SideWinder Force Feedback Pro adapter (CAFE:4004) uses the normal Windows DirectInput and Linux evdev paths without requiring the unlisted-device option. It receives a SideWinder-class 100% recommended output ceiling.
• Raw PicoWinder HID/PID remains gated. The adapter's exact raw report map has not been validated by FFB-Bridge, so Windows stays on DirectInput. macOS actuation is also still unvalidated.

1.3.2-beta.4

• Aircraft impacts produce one bounded physical cue. An MSFS CRASH FLAG or X-Plane has_crashed rising edge raises the existing touchdown one-shot to a stronger profile-scaled thump. The cue fires once per crash edge and remains bounded by Master gain, arm ramp, texture limits, capability resolution, motion Safety Stop, and Disarm.
• Simulator/runtime failure still forces the device off. Disconnect, stale telemetry, provider failure, and application failure remain on the existing ownership-release and physical STOPALL path.
• Seven mechanically back-driven starters now enable restrained AP Follow. C172 G1000, TBM 930, King Air 350i, 747-8, XCub, DA40, and 737-800 start at 10% movement authority with a 20% follow spring. A320 and aircraft whose pilot controls are not mechanically back-driven remain disabled.

1.3.2-beta.3

• Autopilot Follow authority is selectable from 0–50%. The value remains profile-owned and hardware output limits remain authoritative. The Tuning page warns that motor-driven movement is visible to the simulator as pilot input and that higher values can oppose or disconnect the autopilot.
• The 737 shaker has an add-on-safe fallback. Standard simulator stall warning and native shaker telemetry remain primary. When those values are silent, a shaker-capable airborne 737 profile may use valid angle of attack at 88% of valid reported stall AoA. The fallback is silent on the ground, below 40 KIAS, and whenever the envelope is unavailable.

1.3.2-beta.2

• All 13 built-in starters use the complete 1.3.2 profile surface. Manual-reversible controls receive bounded airspeed-gated rotation, surface-follow, and sideslip cues. Hydraulic transports, the A320 passive sidestick, and rotorcraft explicitly suppress sources their controls do not transmit.
• 737 and 747 yoke feel is stronger and axis-specific. Their artificial feel, damping, pitch/roll load balance, and bounded flap, gear, and spoiler cues were retuned for clearer transport behavior while preserving the separate device envelope.
• Configuration cues use actual flap telemetry. Every built-in starter now explicitly chooses flap-motion and flap-arrival behavior.
• AS 33 vertical-wind feel is progressive. The existing source is enabled with a 1.5–15 ft/s envelope and a conservative profile gain.
• Hardware-dependent mechanics remain opt-in. Friction, inertia, and progressive stops still require device- and installation-specific tuning. This candidate adds no 1.4 force engine, schema, simulator, or Aircraft Feel Check work.

1.3.2-beta.1

FFB-Bridge 1.3.2-beta.1 is a focused MOZA AY210 lifecycle and roll-authority test build.

• Flight Check no longer retriggers the AY210 reference movement. FFB-Bridge keeps the AY210 firmware output gates in their ready state during an owned session instead of cycling them at every Arm and Disarm boundary.
• AY210 Disarm still commands absolute force-off. The game force-feedback coefficient and PID gain go to zero before STOPALL, autonomous MOZA effects remain neutral, and repeated Disarm requests reissue the physical stop.
• Routine AY210 cleanup avoids whole-device PID Reset. FFB-Bridge frees only its own effect blocks before recreating the fresh post-STOPALL table. AB6, AB9, and every non-MOZA device retain their existing lifecycle policy.
• MOZA settings still restore exactly. Clean exit restores the journaled 13 firmware values; ownership loss retains the existing fail-closed output inhibition.
• The AY210 automatic recommendation is now 35% pitch and 50% roll output. The independent ceilings reflect the yoke's different linear-pitch and rotary-roll mechanisms. Explicitly saved user limits remain unchanged; existing installations may therefore continue to show 35% Roll until the user selects Reset to automatic or sets Roll to 50%.
• Autopilot Follow no longer softens a firm aircraft profile. Its stiffness setting may increase the normal spring hold, but engaging Follow cannot replace a 737/747-class spring with the former 25% default.

1.3.1

FFB-Bridge 1.3.1 brings the complete public 1.3.1 beta line to Stable: one place for MOZA hardware control, more accurate simulator-driven feel, clearer Flight Check diagnostics, and an absolute force-off boundary whenever the app is disarmed. Existing profiles retain their current feel; the new vertical-wind effect remains off until a user enables it.

• Disarmed now means zero force on every supported device. Disarm sets device gain to zero before STOPALL and does not replay a centering spring, damper, friction, inertia, soft stop, constant force, periodic effect, or firmware autocenter. Every backend opens at zero gain plus STOPALL and shutdown retains zero gain.
• Repeated Disarm is a physical command, not only a UI state change. Every request reissues zero gain plus STOPALL even when software already reports Disarmed. Periodic state reassertion remains silent until explicit Arm. A rejected physical STOPALL is reported instead of being presented as a successful stop.
• MOZA bases use both independent firmware output gates. Managed takeover disables force feedback and asserts torque-output inhibit before changing local damping or other coefficients. Explicit Arm enables the gates; Disarm, startup failure, and shutdown inhibit them again.
• MOZA hardware control is gathered in one place. Hardware → MOZA owns the selected base's pitch, roll, and texture limits, health telemetry, managed 13-setting session, exact shutdown restore, and isolated damping, friction, inertia, and progressive-stop tests. Aircraft-specific feel remains under Tuning. AY210 and AB9 isolated mechanics are verified on real Windows hardware.
• Force trim can use a separate Windows USB grip. A Warthog or similar controller can provide FTR and its optional toggle while the FFB base remains the actuator. Beep trim uses the same guided assignment flow. Nothing is assigned automatically.
• Aerodynamic loading and autopilot follow are more accurate. Pitch and roll loading can be balanced independently, AP-follow bank direction is corrected, Flight Check can use an unmistakable 50% isolated bench motion, and live simulator AP-follow has a hard 8% authority ceiling.
• Flap and engine feel follow more of the real aircraft. Actual flap travel drives movement, drag, buffet, and a separate arrival bump. RPM, combustion, vibration, and reverse aggregate every installed engine—MSFS engines 1–4 and X-Plane engines 0–7.
• Stall and overspeed cues can build progressively. Valid AoA, icing, stall reference, and dynamic redline data refine onset while established simulator warnings remain the fallback. Actual gear and left/right spoiler deployment, measured wheel skid, native shaker intensity, and surface condition improve their matching cues when available.
• Slew and optional telemetry fail safely. MSFS Slew is a hard force-neutral state. Each optional MSFS value is isolated; X-Plane unsupported values keep the established fallback; aircraft changes clear cached optional data. X-Plane 12 also contributes native rotor blade-slap evidence.
• Glider lift and sink can be felt without changing existing profiles. The default-off Vertical wind vibration channel uses aircraft-point environmental wind in MSFS and X-Plane 12, never aircraft climb/descent rate. It remains bounded, silent on the ground, while paused or slewing, and silent when the telemetry is unavailable.
• Windows installer builds verify startup resources. The build scripts clean RID-specific Avalonia intermediates and reject a desktop DLL missing its compiled-XAML loader output instead of packaging a build that exits before its first window appears.
• Profiles and normal armed feel are unchanged. This stable roll-up changes no existing profile automatically and includes no Pro-only simulator, licensing, model, or UI payload.

1.3.1-beta.13

This beta adds a dedicated vertical-wind vibration channel for gliders and thermal soaring while preserving the existing airspeed-loaded control force. Existing Beta-channel installations will be offered this build through the normal consent-first update check.

• Updrafts and downdrafts can now be felt through the pitch axis. The new Vertical wind vibration control increases the turbulence texture as simulator-reported vertical environmental wind strengthens. It uses actual wind at the aircraft rather than aircraft climb or descent rate.
• MSFS and X-Plane 12 both feed the new channel. MSFS uses its vertical aircraft-wind value through an isolated optional SimConnect definition; X-Plane 12 uses the aircraft-point vertical-wind dataref. Missing or invalid telemetry leaves the effect silent instead of inventing a cue.
• The new path stays inside the existing safety envelope. It is silent on the ground, while paused, and while slewing. Vertical-wind and established turbulence evidence combine without stacking into excessive vibration; hardware texture limits, Master gain, arm ramp, watchdog, Disarm, and Safety Stop remain authoritative.
• Existing profiles keep their current feel. Vertical-wind vibration is default-off and no built-in aircraft profile is retuned automatically. Profiles that opt into the new setting require FFB-Bridge 1.3.1-beta.13 or newer.
• Aircraft changes cannot reuse stale wind. X-Plane clears the optional vertical-wind sample when aircraft identity changes, matching the existing capability-reset policy.

1.3.1-beta.12

This beta retains the complete MOZA hardware-control and simulator-telemetry work from beta.9, then makes hardware verification clearer and force shutdown more trustworthy. Existing beta-channel installations will be offered this build through the normal consent-first update check.

• The isolated AP-follow check is now unmistakable. Flight Check can move the control through up to 50% of travel with a firmer spring while no simulator is connected, making pitch and roll direction easy to verify on a bench. Live simulator AP-follow remains capped at 8% so MSFS cannot mistake large force-feedback movement for pilot input.
• AP-follow authority now has a true hard ceiling. Flight-director demand is normalized before authority is applied, so unusually large pitch or bank commands cannot multiply past the stated live limit.
• Disarm and Safety Stop confirm the physical stop command. Completion now requires the force-feedback device to accept STOPALL. An already-disarmed device receives the command again so retained centering springs are halted; a rejected hardware stop is reported as a failure instead of success.
• The 1.3.1 feature set is now frozen. Further 1.3.1 betas will be limited to targeted fixes for issues found during field testing. Normal aircraft profiles, MOZA hardware tuning, and the live AP-follow safety limit are otherwise unchanged from beta.9.

1.3.1-beta.9

This public beta combines the completed beta.8 MOZA/feedback work with a new capability-safe simulator telemetry pass. Existing beta-channel installations will be offered this build through the normal consent-first update check.

• MOZA hardware control is gathered in one place. Hardware → MOZA owns the base output envelope, health telemetry, and isolated damping, friction, inertia, and progressive-stop tests. Aircraft-specific values remain under Tuning → Aircraft control feel. AY210 and AB9 isolated mechanics have been verified on real Windows hardware.
• Force trim can use a separate Windows USB grip. A Warthog or similar controller can supply FTR and its optional toggle while the FFB base remains the actuator. AP-follow roll direction is corrected, and elevator/aileron aerodynamic loading can be balanced independently.
• Flap and engine feel follow more of the real aircraft. Actual flap travel drives movement, drag, and buffet, with a separate arrival bump. The default X-Plane C172 flap buffet is gentler. RPM, combustion, vibration, and reverse now aggregate every installed engine—MSFS engines 1–4 and X-Plane 0–7.
• Stall and overspeed cues can build progressively. Valid AoA plus a plausible stall reference ramps stall buffet before the binary warning; structural ice conservatively advances that envelope. A valid dynamic redline/Vne ramps overspeed buffet. Existing warnings remain the fallback.
• Configuration effects use actual state where available. Real gear and left/right spoiler deployment drive sustained drag/buffet. Measured wheel skid drives brake shudder so ordinary non-skidding braking stays smoother. MSFS native shaker intensity and surface condition are also used.
• X-Plane helicopters gain native blade-slap input. The XP12 blade-slap ratio supplements the existing two-per-rev model.
• MSFS Slew mode is a hard force-neutral state. Constants, springs, textures, one-shots, and device-local condition metadata all drop to zero and recover when slew ends.
• Optional telemetry cannot take down the core stream. Each new MSFS value has an isolated definition, X-Plane zero-only unsupported signals retain the legacy fallback, and cached values clear on aircraft change.

1.3.0

FFB-Bridge 1.3 is the largest release so far: first-class helicopter force feedback, managed MOZA hardware sessions, broader native device support, a self-contained Flight Check, safer startup diagnostics, and a much clearer day-to-day tuning workflow. It is a stable rollup of the complete 1.3 beta and RC series.

• Helicopters are now a first-class control system. Rotorcraft profiles combine cyclic damping, force-trim hold and release, an optional pedestal- style force-trim on/off control, continuous beep trim, AFCS follow-up trim, ETL and VRS cues, two-per-rev vibration, retreating-blade-stall buffet, skid scrape, and rotor rumble. Bell 206, H125, and generic helicopter starters cover the main trim-system families.
• Every helicopter control uses a clear guided setup. FTR and its optional on/off control use the same press-and-release modal. Beep trim asks for one physical four-way control, listens to both device formats, then saves either one POV hat or four buttons. The tested AY210 reports its hat-shaped switch as four ordinary buttons. Input preparation is shown separately with no misleading moving bar, and cancel or failed verification restores the previous assignment.
• MOZA AB6, AB9, and AY210 work as managed FFB-Bridge devices on Windows, Linux, and Apple Silicon macOS. FFB-Bridge snapshots the settings it owns, prepares and monitors a known force-feedback session, prevents stored device effects from fighting the flight model, and restores the exact prior values on a normal exit. MOZA firmware 1.1.4.6 or newer is required. MOZA Cockpit remains useful for firmware, lighting, and maintenance, but should be closed while FFB-Bridge is running.
• Native input and force paths reduce latency and remove normal-use utility dependencies. Supported SideWinder and MOZA hardware use raw HID/PID where validated, with platform-appropriate fallbacks. Held inputs such as beep trim remain continuous, and each device receives the spring, damping, friction, periodic, and one-shot rendering policy validated for its family.
• Flight Check is now the canonical hardware bench test. Direction, centering, damping, friction, trim, autopilot follow, airplane flight, and helicopter flight run from known temporary settings and restore the user's previous state afterward. Safety Stop exercises the real gain-zero, STOPALL, latch, alert, and explicit re-arm path without asking the user to provoke a runaway control.
• Hardware and aircraft feel are separated cleanly. Hardware owns device selection, the physical output ceiling, renderer/backend choice, polarity, calibration, and control assignments. Tuning starts with Profile and Master gain, then one Advanced switch and the Control system selector. Advanced reveals all relevant effect groups in one continuous page, and zero is the single off state for gain controls.
• Profiles start sensibly and follow the aircraft. A previous profile is restored when available; a fresh install starts on the Cessna 172; reported aircraft identity can then select a closer starter or a saved aircraft binding. The compact profile library adds search, sorting, import, duplicate, rename, share, and protected read-only starters.
• Startup health checks are automatic but stay out of the way. Healthy launches remain silent. Actionable device or permission failures produce a non-blocking Review-and-fix banner that opens populated Support rows. Linux verifies the active event node and complete udev coverage, including MOZA TTY policy channels; after a successful rule repair, the app recommends a restart so it can reopen the joystick with the new permissions.
• The force model is broader and more device-aware. 1.3 adds native control friction, glider airspeed feel, simulator-driven flap motion and buffet, improved Safety Stop recovery, device-normalized output ceilings, expanded roll and autopilot behavior, and a conditional 200 Hz loop for renderers that need host-side waveform or device-coupled updates.
• Hardware coverage is wider. Alongside the Microsoft SideWinder Force Feedback 2 and MOZA family, 1.3 includes the listed Logitech G940, Force 3D Pro, and WingMan Force 3D paths plus Brunner FFB-G support. Experimental unlisted joystick opt-in remains conservative and excludes wheels, gamepads, and unsuitable single-axis devices.
• Updates and shutdown are safer. Windows and packaged Linux installs can download, verify, install, and restart after explicit consent; Windows also verifies the expected Authenticode publisher. macOS opens the signed current download. Managed device restoration completes before exit, and Linux tray observers are disposed before dispatcher shutdown.

Supported simulators are Microsoft Flight Simulator 2024/2020 on Windows and Linux, and X-Plane 11/12 on Windows, Linux, and Apple Silicon macOS. Existing profiles and hardware preferences migrate forward; control assignments remain explicit and simulator keybindings are never changed.

1.3.0-rc1

The first 1.3 release candidate completes the usability and startup-safety work developed after beta.39. It is intended for final cross-platform field validation before 1.3.0 stable.

• Startup device problems are detected automatically on Linux, Windows, and macOS. FFB-Bridge quietly checks the physical force-feedback device when it opens. Linux also checks a supported event-node's permissions and installed udev coverage, including the additional MOZA interfaces required for managed setup. Passing checks stay out of the way.
• Actionable startup failures are now impossible to overlook. A red modal identifies the failed checks and leads directly to Support → Health checks. It can be dismissed for an intentionally disconnected device, but automatic force arming remains blocked until the warning has been reviewed.
• Tuning is organized around the normal adjustment sequence. Profile and Master gain come first, followed by a standard Advanced switch and a compact, explained Control system selector. Enabling Advanced reveals and expands every available tuning section into one continuous page for scrolling.
• Tuning controls are clearer and more consistent. Slider labels, ranges, help, exact-value entry, dirty state, and reset behavior now come from one shared catalog. Redundant effect-enable checkboxes are gone: setting the relevant gain or authority to zero is the single off state, including safe normalization of older saved profiles.
• Profiles and hardware setup take less space and give better feedback. Profile search, sorting, and import are consolidated; list values retain a scrollbar-safe gutter. Dedicated beep-trim POV setup now acknowledges each detected direction and keeps the previous mapping if learning fails.
• MOZA AB6, AB9, and AY210 are confirmed supported hardware. Managed configuration ownership works across Windows, Linux, and macOS. Narrower effect-quality evidence remains tied to the exact device and backend that produced it.
• The desktop window behaves like a normal native application again. It can be resized and maximized, restores screen-safe geometry, and supports a smaller practical minimum size without clipping the main workflows.
• Linux shutdown no longer races the desktop D-Bus dispatcher. Tray icons and their observers are disposed while Avalonia's UI dispatcher is still active, preventing the post-cleanup TaskCanceledException previously recorded after MOZA restoration and device shutdown had already completed.

1.3.0-beta.39

Cross-platform MOZA setup and update-channel correctness rollup. This public build brings together everything added after beta.35.

• Linux and macOS now establish the MOZA parameter session correctly. The shared serial transport performs the required motor identity exchange before reading settings, matching the proven Windows transaction and unlocking replies from a freshly connected AB6, AB9, or AY210.

• Linux finds and permits every required MOZA interface. Physical sysfs topology now pairs the correct /dev/ttyACM* policy channel by exact USB identity. The generated rule covers the input and MOZA tty nodes; the Support health check flags older event-only rules for repair. No custom hidraw rule is installed.

• Device pairing is stricter and safer. Absent Linux/macOS models are skipped before firmware ownership, macOS deduplicates the cu.* and tty.* names for one interface, and ambiguous multi-base configurations are refused instead of pairing the wrong HID and policy channels.

• AB9 and AY210 are hardware-validated on Linux and macOS. Both bases now complete managed setup, raw/kernel force output, input reads, and exact original-setting restoration on clean exit.

• Linux setup controls remain reachable. The expanded MOZA rule preview scrolls within the dialog so Cancel and Install rule stay visible.

• Stable and Beta checks now name the release they actually checked. A newer local beta on the Stable channel reports both the installed version and latest stable version instead of calling the installed beta “up to date” without context.

• Failed checks no longer claim the app is current. Network errors and invalid manifests now show a retry message, while startup checks remain silent.

• Channel boundaries are enforced. Stable rejects a prerelease manifest, and switching channels clears or ignores results from the previous channel.

1.3.0-beta.35

This release brings together everything added from beta.19 through beta.35. It is a major hardware-support and usability update, especially for MOZA owners.

• MOZA AB6, AB9, and AY210 no longer require MOZA Cockpit for normal use. FFB-Bridge prepares the base automatically when it starts, keeps the required device settings correct while you fly, and works independently on Windows, Linux, and macOS.
• Your MOZA settings are returned when FFB-Bridge closes. The app preserves the base's original force-feedback state, mode, strength, and built-in effect settings, then restores and verifies them after a normal exit. If a session is interrupted, the saved state remains available for recovery next time.
• MOZA setup and restoration are now visible. Clear startup and shutdown messages show when the base is being prepared or restored. On Windows, FFB-Bridge also detects MOZA Cockpit and asks you to close it before the two programs can compete for the device.
• Raw HID/PID is now the preferred Windows path for the MOZA AB6, AB9, and AY210. Axes, buttons, and hats are read directly and continuously, fixing delayed or buffered held controls such as beep trim. DirectInput remains available as a compatibility option.
• Nothing is assigned without permission. FTR, beep trim, toggle controls, and dedicated hats remain unassigned until you enable or learn them under Hardware. FFB-Bridge never changes simulator keybindings.
• MOZA damping and friction feel stronger and more consistent. AB6 and AB9 use the response best suited to their shared hardware family, while AY210 uses its validated native effects. Stored background effects are prevented from stacking with the forces generated by FFB-Bridge.
• Control friction is now available for supported sticks and yokes. It adds adjustable breakout and sliding resistance without pulling the controls toward centre. Profiles keep it off by default, so existing setups do not change unexpectedly.
• Flight Check is now the one place to validate hardware feel. Its damping and friction checks use known temporary test settings, do not depend on the active aircraft profile, and restore the user's previous state afterward. The friction check now runs at full test strength so missing or weak hardware response is easy to identify.
• SideWinder FFB2 now prefers its complete raw HID/PID path on Windows. DirectInput remains available automatically or by explicit choice if a compatibility fallback is needed.
• Hardware and aircraft tuning are easier to understand. Hardware contains device-specific limits and calibration; Tuning contains aircraft feel. The Tuning page now starts with the essential controls, hides expert options behind one Advanced switch, shows only controls relevant to the selected aircraft type, and removes the obsolete steady-force smoothing option.
• Safety Stop is much less likely to interrupt aggressive normal flying. It now requires substantially stronger and longer evidence of runaway motion before it trips. When a real trip is detected, output is still stopped and must be re-armed explicitly.
• Updates can now complete inside FFB-Bridge on Windows and Linux. After explicit approval, Windows and packaged Linux AppImage installs download the new build, verify its exact length and SHA-256, install it for the current user, and restart the verified copy. Windows also requires the expected publisher signature. macOS continues to check automatically and opens the current release notes and signed download.

1.3.0-beta.19

Native control friction for sticks and yokes, added from pilot feedback and validated on real MOZA hardware.

• New optional Control friction effect. It models the breakout and sliding resistance of bearings, pushrods and control cables without pulling toward centre or growing with movement speed. It is separate from aircraft-rate damping and helicopter cyclic damping.
• Control friction is rendered by the device, not a delayed software loop. Unsupported hardware remains safely inert. It is enabled only on explicitly admitted configurations, including the SideWinder FFB2, Brunner FFB-G, and Windows DirectInput MOZA AB9/AB6 and AY210; Linux uses it only when the active device advertises native FF_FRICTION support.
• A dedicated Tuning card keeps the behavior clear. The feature is off by default and existing profiles remain physically unchanged. On supported hardware, begin around 5–10% and increase only until the linkage feel is convincing.
• Flight Check now tests friction in isolation. It uses the current Tuning strength even when live-flight friction is disabled, asks you to move the stick or yoke through pitch, roll, diagonals and reversals, and refuses to run a misleading 0% test.
• Dashboard and profile handling understand the new state. The Stick feel detail reports configured/ready friction, saved profiles migrate additively, and older profiles default to friction off.
• Fixed the DirectInput per-axis condition encoding. Friction now uses the unrotated Cartesian form expected for two axis conditions. On the AY210, two repeated maximum-output tests produced a definite increase in resistance; the effect lifecycle also completed cleanly through stop and reset.

1.3.0-beta.18

Safety-stop recovery and simulator-driven flap feedback, validated on the MOZA AY210.

• Safety stops are now unmistakable and simple to recover from. The Arm control turns into a red SAFETY STOP. One explicit click re-arms, and force output returns gradually over two seconds. The cut never clears or restores forces silently.
• An optional spoken warning now accompanies the stop. The bundled offline alert says “Safety stop. Forces disabled.” and makes no network request. It can be disabled under Settings → Session.
• Flight Check can safely exercise the real emergency path. Test safety stop runs the same device-gain-zero, STOPALL, latch, status, alert, and explicit-rearm path without violent stick movement or a running simulator.
• Normal aggressive inputs are less likely to false-trip the guard. The detector now requires three rapid reversals while retaining the latched cut for sustained runaway motion.
• Fixed an arm/disarm notification race that could hard-lock the app. A stale UI callback can no longer fight the runtime during fast reconnects with Auto-arm and Disarm-on-exit enabled.
• Flap feedback now follows the physical surfaces reported by the simulator. MSFS trailing-edge and X-Plane flap deployment telemetry drive movement vibration only while the surfaces are travelling. Sustained flap buffet and drag follow actual deployment, while the handle-detent pulse remains a separate one-shot.
• X-Plane now uses each aircraft’s reported flap-detent count. This preserves aircraft-specific handles such as the stock 737’s eight positions.
• Diagnostics remains responsive during bursts of simulator events. New log rows still appear immediately, while summaries and triage cards are coalesced instead of rebuilding the full log for every event.

1.3.0-beta.17

Major force-safety and helicopter-damping update, driven by real AB9, AY210, SideWinder FFB2, and Logitech reports.

If you use a modern direct-drive base, begin at the new recommended physical-strength limit, keep hands clear while arming, and keep the base's power switch or emergency stop within reach. Choosing Advanced and raising the limit toward 100% deliberately permits the base's full advertised force.

• SideWinder FFB2 cyclic damping no longer runs through a delayed software feedback loop on Windows. Both the default raw HID/PID backend and the DirectInput backend now use the stick's native two-axis Damper condition. This fixes the violent rotating shake / periodic bumps some FFB2 users felt from tiny cyclic movements. Both Windows backends passed the complete helicopter Flight Check and repeated arm/disarm cycles on real hardware.
• Listed Logitech sticks get the same device-native damping fix on DirectInput. Flight System G940, Force 3D Pro, and WingMan Force 3D now use their native Damper condition instead of the host-side loop, addressing reports of heavy bumps unless cyclic strength was reduced to very low values.
• MOZA keeps its proven cross-platform damper. AB6, AB9, and AY210 continue using the validated 200 Hz synthesized damper on Windows, Linux, and macOS; they are not silently switched to an unproven native DirectInput effect. A polarity-domain bug was also fixed so changing a spring-centre sign can never reverse damping into a force that accelerates the stick in the direction it is already moving.
• New per-device Maximum physical strength control. Hardware → Effects now shows the actual fraction of the base's command range. Recommended starting ceilings are 35% for AB6/AB9/AY210/Brunner, 20% for an unlisted device, and 100% for the SideWinder FFB2 and listed legacy Logitech sticks.
• The recommendation is not a lockout. A clear strong-base warning unlocks the slider through 100% raw hardware output, so experienced users can use the full capability of their base deliberately. Saved overrides store the actual physical percentage, so a future change to a recommendation cannot silently make an existing setup stronger.
• The physical ceiling covers every bridge output path and every backend. Live flight, Flight Check, Mock Sim, raw HID/PID, DirectInput, evdev, IOHID, software-blended effects, native springs/dampers, firmware periodics, one-shots, and AP/trim backdrive all pass through the same final device limit. Spring centre and deadband geometry remain unchanged.
• Safer force transitions. Arming begins at zero and ramps toward the selected ceiling. Raising the ceiling ramps in; lowering it takes effect immediately. Directed-force reversals and continuous-effect onsets are rate-limited to avoid abrupt full-scale steps.
• Master Gain now does what it says during bench tests. Flight Check steps, both guided finales, Mock Sim, the direct direction/polarity test, cyclic damping, ETL/VRS, buffets, springs, and one-shots now inherit the active Master Gain. A 10% bench check is much gentler than 100%.
• Dashboard Stick feel Off now includes helicopter feel. Muting the group disables cyclic damping and force trim as well as the existing fixed-wing stick-feel effects; turning it back on restores them.
• Changing polarity or swapping axes can no longer happen under power. If forces are armed, the bridge completes a device disarm before applying the new motor mapping, stops any running calibration test, and requires an explicit re-arm.
• Runaway-motion protection is harder to bypass. The device-wide cut now covers both render pipelines and irregular two-axis motion, zeros device gain before STOPALL, clears replay state, stays latched through watchdog/disarm cleanup, and reports the retained reason on the Dashboard. Only an explicit Arm clears it.
• Repeated FFB2 tests no longer churn the firmware effect table. Raw HID/PID and DirectInput keep their Constant/Spring/Damper allocations warm across re-arms, fixing corrupt allocation replies and effects going dead after repeated Flight Check cycles. MOZA retains its separately validated recreate-after-STOPALL lifecycle.
• Removed the recurring false native-crash popup. An armed-session marker by itself can also mean Task Manager, debugger termination, power loss, or a development rebuild, so it no longer fabricates a pid.dll access-violation report or switches effect rendering. Genuine managed/SEH crash reports still trigger recovery, and obsolete inferred reports are removed automatically without deleting real reports.
• Regression coverage expanded to 1,582 tests. On Windows, 1,581 pass; the only local failure is the pre-existing Proton/Linux path-separator assertion that passes on the Ubuntu release runner. The new coverage pins output ceilings, arm/increase ramps, immediate reductions, direction slew, native/synth damping selection, override migration, Master Gain propagation, V3 parity, and validation-tool persistence.

1.3.0-beta.14

Safety release — please update, especially on MOZA and other direct-drive bases.

• New safety cut for runaway oscillation. If the stick ever builds into a violent, sustained back-and-forth oscillation (reported on an AB9 with the base's own damping turned off), the bridge now detects it from the stick's own motion and stops all force output within about a second, with a message telling you what to adjust. Disarm and re-arm to resume. Normal flying, firm corrections, and the natural brief overshoot when you release the stick do not trigger it.
• MOZA setup correction: set the Cockpit Damper to about 10%, not 0. Our earlier advice said to zero it — that was wrong and is what made the oscillation possible. The damper resists motion, not position, so it doesn't fight the bridge's centering; on a strong direct-drive base it's what makes the stick settle instead of bouncing. Spring, Inertia and Friction stay at 0. The in-app reminder, Support health check, and setup guide are updated.
• Fixed clean quits sometimes being misread at the next start as a crash (the "previous session did not shut down cleanly" message switching you to software rendering). If that happened to you: Hardware → Effects → Effect rendering → Hardware puts you back.

1.3.0-beta.13

• MOZA bases (AB9, AB6, AY210) now centre using the base's own hardware spring directly — firmer, smoother centering that settles cleanly when you let go. This also fixes an issue where the previous synthesized centering could shake back and forth on a strong direct-drive base with no hand on the stick.
• MOZA setup: in MOZA Cockpit, set Spring, Damper, Inertia and Friction all to 0 and Overall Force Feedback Intensity to 100% (the in-app reminder and setup guide now say this). A lower intensity quietly scales every force down — including fine cues like beep trim, which can otherwise feel dead for the first moment on a stiff base until it catches up.
• Flight Check now has one clear "Centering spring" test instead of the old hardware/synthesized pair.
• Fixed a false "previous session didn't shut down cleanly" warning that could switch you to software rendering after a perfectly normal quit (most likely on MOZA hardware). Clean exits are now recognised reliably, so hardware effects stay on unless there's a real crash.

1.3.0-beta.11

• Swap your force-feedback stick and the bridge now notices, and resets the device-specific settings (backend, effect rendering) to the recommended defaults for the new stick — with a one-time notice of what changed. If nothing needed changing, you won't see it.

1.3.0-beta.10

• MOZA AY210 yoke: full force feedback now confirmed on Windows, Linux, and macOS (raw HID/PID / evdev). No setup difference from other MOZA bases.

1.3.0-beta.9

• Fixed a crash at quit while armed (seen on a Logitech Force 3D Pro, applies to all sticks): quitting could briefly restart the sim-connection logic mid-shutdown, racing device teardown. This was also what triggered the "previous session did not shut down cleanly" recovery message and the switch to software rendering for some users. If that happened to you: Hardware → Effects → "Test hardware effects" restores hardware mode.

1.3.0-beta.8

• SideWinder FFB2 on the DirectInput backend: fixed forces going wrong or dead during Flight Check. A MOZA-specific recovery behaviour was accidentally applied to every DirectInput stick, resetting the FFB2's effect memory on every re-arm — which its driver stack does not tolerate. Now correctly MOZA-only.
• If you inverted any axis or centering polarity while chasing this, undo it after updating — with the fix your forces are correct again and the inversion would reverse them.

1.3.0-beta.7

Flight Check fixes from FFB2 tester feedback:

• Force trim hold strength works again. A beta.2 regression made hold-strength changes silently never reach the stick once the hold was firm — adjusting the slider now changes the felt hold.
• The force trim Flight Check step now says what actually happens: the stick starts LOOSE and engages on first FTR use.
• SideWinder FFB2: gentler on the stick's control pipe (one input read per tick instead of three) and a more patient effect-table recovery on re-arm — addresses tests going silent until an app restart.
• Learning the force-trim on/off switch onto the same button as FTR no longer makes FTR appear dead.

If your Flight Check direction tests seem reversed, please UNDO any axis inversion you set as a workaround and send a support bundle (Support → Diagnostics) — we're chasing that report and the inversion also reverses live flying.

1.3.0-beta.6

• Fixed the broken layout of the update settings in Settings → General (the channel picker crushed the text into a one-word-wide column).
• First release published through the new auto-update channel — if you're reading this from the in-app "What's new" link, it worked.

1.3.0-beta.5

• Auto-update. FFB-Bridge can now keep itself current: it checks for a new version at startup (on by default — turn it off in Settings → General), shows what's new, and on Windows downloads the signed installer, verifies it, and updates in place. Choose the stable or beta channel in Settings. The check fetches a small public version file and sends nothing about you or your system.
• Release notes now live at ffb-bridge.com/releases.

1.3.0-beta.4

Helicopter force-trim update, driven by beta-tester feedback:

• Firm hold at the trim point. The force-trim hold now behaves like the real magnetic brake — it resists immediately at the trimmed position instead of easing in (the reported "null zone" is gone).
• Force trim on/off button. Optional second learned button that works like the pedestal FORCE TRIM switch: press once and the cyclic goes fully loose (hover work), press again and the brake grabs at the current stick position.
• Beep trim. Nudge the stick's hat switch to walk the trimmed position slowly — no setup needed; hatless sticks can learn four buttons instead.
• Follow-up trim (AFCS types). Optional per-profile behaviour: below 40 knots, steady stick pressure slowly re-trims toward the stick, as the EC135/H225 family really does.
• Every helicopter gets the right trim style automatically. Load any helicopter and the bridge picks classic force trim, AFCS trim, or no-force-trim-with-light-springs based on the real aircraft — plus a new Trim style picker in Tuning → Helicopter.
• MOZA AB9: fixed forces going dead after a disarm/re-arm cycle.

1.3.0-beta.1

The helicopter beta (testers wanted — see the community thread):

• Complete helicopter support: cyclic damper, force trim (FTR), ETL shudder, VRS buffet, and rotor feel pack 2 (two-per-rev, retreating blade stall, skid scrape, rotor rumble).
• Windows raw-HID input read: FTR and the damper now work on the default SideWinder FFB2 backend — no more DirectInput workaround.
• 200 Hz feel tick: the damper is stable at full strength on light sticks.
• Six new aircraft starters: XCub, DA40, P-51D, 737-800, Extra 300, and the Airbus H125.
• Glider support: proper speed-based control loading (≈ V²) and an AS 33 starter profile.
• Hardware settings moved to their own page; Flight Check covers far more (roll checks, autopilot, centering modes, helicopter steps).
• MSFS 2020: the bridge now finds and fixes the 2020 SimConnect config too.
• Brunner FFB-G joins the supported-device registry.

1.2.1

• MOZA AB9/AB6 hotfix: synthesised centering spring — MOZA bases accept but don't render the DirectInput spring effect, so the bridge now reads the stick position and renders centering itself. Restores self-centering and trim's hands-off hold on MOZA hardware.

1.2.0

• Flight Check: a per-effect diagnostic page — confirm every force the bridge makes, one guided step at a time, no sim needed.
• Elevator/aileron trim sign closure (live-MSFS validated) and a one-shot polarity reset for configs that carried workaround inversions.
• MOZA AY210 yoke base joins the supported-device registry.