Spoilers — Roll and Speedbrake

Each wing carries a bank of six spoiler panels, numbered 1 to 6 from the fuselage outward along the rear upper surface. They are the chapter's clearest example of one piece of aluminium doing several unrelated jobs: the same panel that helps the ailerons bank the aircraft (roll assist) also stands up symmetrically with its partner across the wing to add drag in flight (speedbrake), and dumps lift wholesale on touchdown (ground spoiler). The physical motion is always the same — the panel rises to spoil lift — but it serves three completely different control logics. This article enlarges the first two. The third, the automatic lift-dumping logic at landing and rejected take-off, is a system in its own right and is deferred to Ground Spoilers and Function Allocation.

Two anchors first, because both run against instinct and reframe the failure cases that follow:

[!warning]- Loss of electrical control and loss of hydraulic pressure leave a spoiler in opposite resting positions.

The reflex is "any failure folds the panel away." Only half of that is true. Lose the electrical signal and the panel is actively retracted to zero — the servovalve's biased zero drives it down while hydraulic muscle is still available. Lose the hydraulic supply and the panel does the opposite: it freezes at the deflection it had, then is pushed down only by the airflow, slowly and passively. Same word "failure", two opposite behaviours — and which one you are looking at decides whether the panel is parked safely or hanging in the breeze.

[!warning]- A failed spoiler is not taken over by another computer — it stays lost.

Per the chapter's master logic, every other surface has a computer standing by to execute a failed computer's task — except spoiler control. Each panel is driven by exactly one computer with no electrical back-up, so a spoiler fault presents as "these particular panels are permanently unavailable" rather than "another computer picks up the job." This is why spoiler dispatch is framed around which pair is lost, not which computer failed (see Flight Control Fundamentals).

1. The six-panel bank — three jobs, one motion

The maintenance source states the panel count and the job list in one paragraph. Per AMM 27-60-00:

Six spoilers (1 thru 6) are installed on the rear upper surface of each wing. They are used in different configurations for: [...] roll control (spoilers 2 thru 6 and ailerons) [...] speed brake and ground spoiler function (spoilers 1 thru 6) [...].

Read the bracketed panel ranges carefully, because they are the key to the whole article:

• Roll control uses spoilers 2 thru 6 — five panels, with the ailerons. Spoiler 1 does not assist roll.
• Speedbrake and ground spoiler use spoilers 1 thru 6 — all six, including spoiler 1.

So spoiler 1 exists purely as a deceleration and lift-dump panel; it never takes part in banking the aircraft. That single fact explains why, later, spoiler 1 has the shortest travel and a different deflection limit from its neighbours.

FCOM opens its roll description with the surface count and the maximum deflection. Per FCOM DSC-27-10-20:

Two ailerons and five spoilers on each wing control the aircraft about the roll axis. The maximum deflection of the ailerons is 25 °. Ailerons extend when the flaps are extended (aileron droop). The maximum spoiler deflection is 35 °.

The "five spoilers" is the roll count (panels 2–6); the 35° is the maximum a spoiler can reach — and, as §3 shows, that maximum belongs to the roll-assist job, not the speedbrake job. The aileron half of the roll axis is enlarged in Ailerons; this article follows the spoilers.

2. Architecture — hydraulic × computer × function

The single most useful picture of this system is the one that puts each panel's hydraulic supply, its commanding computer, and its function set side by side. The diagram below is built from the FCOM and AMM allocations verified in this article; read each column as one panel.

        FUSELAGE ◄────────────── panels 1..6 ──────────────► WINGTIP
       ┌───────┬───────┬───────┬───────┬───────┬───────┐
 HYD   │   G   │   B   │   B   │   Y   │   G   │   Y   │   Green = 1 & 5
 panel │   1   │   2   │   3   │   4   │   5   │   6   │   Blue  = 2 & 3
       ├───────┼───────┼───────┼───────┼───────┼───────┤   Yellow= 4 & 6
 CPU   │ PRIM3 │ PRIM3 │ SEC 2 │ PRIM2 │ PRIM1 │ SEC 1 │
       └───────┴───────┴───────┴───────┴───────┴───────┘
 roll      —       o       o       o       o       o       (roll: 2-6 only)
 SB        o       o       o       o       o       o       (speedbrake: 1-6)
 GND SPL   o       o       o       o       o       o       (ground: 1-6 → art.23)
 MLA       —       —       —       o       o       o       (load alleviation 4-6)
 travel  50.7mm  69.5mm  69.5mm  69.5mm  69.5mm  69.5mm
 max °    25°SB   30°SB   30°SB   30°SB   30°SB   30°SB    (roll assist up to 35°)

Four things to read off it, each developed below:

1. Hydraulics are strictly paired. Green feeds 1 & 5, Blue feeds 2 & 3, Yellow feeds 4 & 6 — exactly two panels per system per wing. So losing any one hydraulic system removes only two of the six panels on that wing; four remain. This is the root of the spoiler's resilience to a single hydraulic loss (§9).
2. Each panel has its own computer, with PRIM and SEC interleaved. Panels 1 and 2 on PRIM 3, panel 3 on SEC 2, panel 4 on PRIM 2, panel 5 on PRIM 1, panel 6 on SEC 1. Spreading five roll panels across different computers means no single computer or hydraulic loss takes more than one or two of them.
3. Spoiler 1 is the odd one out — no roll, no MLA, shortest travel — the dedicated deceleration panel.
4. Panels 4–6 are the busiest — roll, speedbrake, ground spoiler and Manoeuvre Load Alleviation (MLA). The MLA role is developed in Load Alleviation.

2.1 The hydraulic allocation

Per AMM 27-64-00:

Hydraulic power from the three aircraft hydraulic systems supplies the six spoilers on each wing. The Green hydraulic system supplies spoilers 1 and 5. The blue system supplies spoilers 2 and 3. The Yellow system supplies spoilers 4 and 6.

These pairings are absolute. Memorise "Green 1 & 5, Blue 2 & 3, Yellow 4 & 6" — every spoiler-related hydraulic-failure consequence falls straight out of it. The three systems are the spoiler's only force source: a spoiler has no electric muscle, so when its hydraulic system is gone the panel cannot be actively driven at all (the basis for the hydraulic-loss "freeze" in §7).

2.2 The control computers

FCOM gives the roll allocation directly. Per FCOM DSC-27-10-20:

The SECs control the N° 3 and 6 spoilers ; the PRIMs control the N° 2, 4 and 5 spoilers.

The exact computer number for each panel is fixed by the aircraft schematic (verified panel by panel): spoilers 1 & 2 → PRIM 3, 3 → SEC 2, 4 → PRIM 2, 5 → PRIM 1, 6 → SEC 1 (per ASM 27-98-31 thru 35). For the speedbrake and ground-spoiler functions the command runs through a PRIM-then-SEC chain (§4), but the actuating computer per panel is still a single one: per AMM 27-60-00, One FCPC or one FCSC controls each servocontrol. That single-computer-per-panel design — with no electrical back-up — is the hardware reason behind the "spoiler is not taken over" rule in the opening callout.

[!warning]- The deflection figures in millimetres and degrees describe different things — do not equate 50.7 mm with a deflection angle.

The millimetre figures are the actuator's controlled stroke from the AMM hardware description; the degree figures are the aerodynamic deflection limits from FCOM. They map onto each other panel by panel but live in different layers. Spoiler 1's 50.7 mm stroke corresponds to a 25° speedbrake limit; the 2–6 stroke of 69.5 mm corresponds to 30° for the speedbrake. And the absolute spoiler maximum is 35°, reached only in the roll-assist job — never as a speedbrake (§3).

3. Roll assist — how the spoilers help the ailerons

A downward-deflected aileron is a poor roll device at the wrong end of the envelope: at high angle of attack the down-going aileron can stall first, so roll authority falls away just when it is needed. The spoiler closes that gap, because it only ever spoils lift — it rises on the wing that needs to drop, killing lift there to deepen the bank. So the A330 banks with ailerons plus spoilers together, and the spoiler's contribution is asymmetric: the panel rises on the down-going wing while the opposite wing's spoilers stay flat.

FCOM has already given the roll allocation — SECs on 3 and 6, PRIMs on 2, 4 and 5 — which spreads the five roll panels across four different computers. That interleaving is the core of the redundancy: any one computer or hydraulic system takes at most one or two roll panels with it, and the survivors still generate a rolling moment.

[!warning]- The same panel deflects further as a roll surface (up to 35°) than as a speedbrake (30°, or 25° for panel 1).

Both numbers are FCOM figures for the same hardware, and they are not a contradiction. Roll is a single-sided, fast, large-authority demand, so the panel is allowed its full 35°. Speedbrake is a symmetric, pure-drag demand where 30° is sufficient and excessive deflection would invite buffet and pitch disturbance, so it is capped lower — and reduced further still in CONF 2/3/FULL (§4). When you see a spoiler beyond 30° on the ECAM F/CTL page, it is doing roll, not speedbrake.

The asymmetric rise-one-wing behaviour is exactly what collides with the speedbrake's symmetric rise-both-wings behaviour when both demands arrive at once — which is the subject of §5.

4. Speedbrake — one lever, symmetric panels

FCOM names the control and the panel set in two short lines. Per FCOM DSC-27-10-20:

The flight crew controls the speedbrakes with the speedbrake control lever. The speedbrakes are spoilers 1 to 6.

The speedbrake uses all six panels — including spoiler 1, which exists for precisely this — and raises them symmetrically on both wings so the drag is pure, with no rolling moment. The maintenance source traces the signal path the lever takes, which FCOM does not. Per AMM 27-60-00:

The speedbrake control lever transmits electrical signals to the associated Flight Control Primary Computers (FCPC). [...] The FCPCs extend the signals and transmit them to the Flight Control Secondary Computers (FCSC). [...] The FCSCs send the signals to the spoiler servocontrols.

The lever talks to the PRIM first, not straight to the panels. That is deliberate: the speedbrake has to respect angle-of-attack, speed and configuration inhibitions (§6), and those are computed in the PRIM. The PRIM decides whether and how far to extend, then extends the order down to the SEC, which drives the servocontrols. This is the surface-level confirmation of the chapter rule that the PRIM owns speedbrake and ground-spoiler control (see EFCS Computer Architecture).

The deflection limits, and their reduction with flap, are given by FCOM. Per FCOM DSC-27-10-20:

The maximum deflection for the spoilers is: • 25 ° for spoiler 1 • 30 ° for spoilers 2 to 6. The maximum deflection for the spoilers is reduced in CONF 2, 3 and FULL.

The deeper the flap setting, the more the speedbrake is held back — because a high-lift wing is already running closer to flow separation, and over-raising the panels there would provoke buffet and a pitch disturbance. The reduction is driven by the flap-configuration signal the slat/flap computers supply to the PRIM.

5. Roll has priority over speedbrake

Panels 2–6 carry both a roll order and a speedbrake order at the same time. Imagine descending with the speedbrake out (both wings raised, say, 25°) and then rolling: the down-going wing's panel is asked for 25° of speedbrake plus a roll increment, which can exceed the deflection achievable in flight. Something has to give, and FCOM is explicit about what. Per FCOM DSC-27-10-20:

For surfaces 2 to 6 (that perform roll and speedbrake functions), the roll function has priority: When the sum of a roll order and a simultaneous speedbrake order on one surface is more than the maximum deflection achievable in flight, the symmetric speedbrake on the other wing is retracted until the difference between the two surfaces is equal to the roll order.

The key is what gets sacrificed. The system does not trim the panel that is already maxed out; it lowers the speedbrake on the opposite wing. Roll is produced by the difference between the two wings' panels, so retracting the opposite wing restores the required difference without asking the saturated panel for more. The cost is real and worth flying with awareness of: while you hold bank, the speedbrake's deceleration is temporarily reduced because the opposite wing's panels have come down. This is why a large rolling manoeuvre and "full speedbrake effect" cannot be expected together.

6. Speedbrake inhibition — five reds and an automatic retraction

A raised speedbrake spoils lift, raises the angle of attack the wing needs, and sheds wake into the tailplane — all of which are exactly wrong when the aircraft is already short of lift or near a protection. So the system actively forbids and auto-retracts the speedbrake in five cases. Per FCOM DSC-27-10-20:

Speedbrake extension is inhibited, if: ‐ MLA is activated, or ‐ Angle-of-attack protection is active, or ‐ Low speed stability is active, or ‐ At least one thrust lever is above MCT position, or ‐ Alpha floor is activated. If an inhibition occurs when the speedbrakes are extended, they automatically retract and stay retracted until the inhibition condition disappears, and the flight crew resets the lever (the speedbrakes can be extended again, 5 s after the lever is reset).

What each red light means to the pilot:

• MLA activated — panels 4–6 have been commandeered for Manoeuvre Load Alleviation (unloading the outer wing); they cannot simultaneously serve as speedbrakes (Load Alleviation).
• Angle-of-attack protection / low-speed stability / alpha floor — all three say the aircraft is at or approaching a low-speed, high-AoA state. Spoiling lift there would push it further toward the stall, so the system retracts the speedbrake to protect your lift margin (Angle-of-Attack Protection, Alpha Floor / TOGA Lock).
• A thrust lever above MCT — you are commanding high thrust (climb or go-around); a speedbrake fighting that thrust is a contradiction, so it is inhibited.

The procedural sting is in the tail: after an inhibition the speedbrake does not pop back out by itself. The crew must reset the lever (push it back in, then out again) and even then it stays in for 5 seconds before it can re-extend. This deliberate re-arming requirement stops the panels from suddenly extending unnoticed once the condition clears.

A separate, symmetric rule governs a failed speedbrake panel. Per FCOM DSC-27-10-20:

When a speedbrake surface on one wing fails, the symmetric speedbrake surface on the other wing is inhibited.

A speedbrake must be symmetric or it produces a rolling moment. If one panel fails (jammed or retracted) and its opposite number kept extending, the lift asymmetry would roll the aircraft. So the system disables the matching good panel as well, accepting the loss of one pair's worth of drag to keep the deceleration symmetric. (Note that the roll function treats panels 4 and 6 as an exception to this pair-disable logic — see §7.)

7. Inside the actuator — biased zero, and the opposite failure modes

Each panel is driven by a single electrohydraulic servocontrol — hydraulic force, electrical command — the panel-level expression of the chapter's control-versus-actuation split. Per AMM 27-64-00:

Spoiler hydraulic actuation is provided by electrohydraulic servocontrol. Each spoiler has an electrohydraulic servocontrol which operates hydraulically and is controlled electrically.

Each servo carries a Linear Variable Differential Transducer (LVDT) that closes the position loop back to the computer. Per AMM 27-64-00:

The spoiler servocontrols get position commands from the flight control primary and secondary computers. Each spoiler servocontrol has a Linear Variable Differential Transducer (LVDT). The LVDTs give information about the position of the spoilers to the associated flight control computer. LVDT faults are detected by the related flight control computer.

The actuator body itself is a uniform design across all six panels, differing only in stroke. Per AMM 27-64-00:

The servocontrols are all the same size but have different lengths of travel. The servocontrol for spoiler 1 has a controlled travel of 50.7 mm (1.9961 in.). The servocontrols for spoilers 2 thru 6 have a controlled travel of 69.5 mm (2.7362 in.). The servocontrols are of a double acting linear type. They have two eye ends one of which is installed to the wing structure, the other is installed to the spoiler.

Three points to extract: the servo is double-acting (it can drive the panel both up and down, so the panel does not rely on a return spring); it is anchored by two eye ends, one to the wing box and one to the panel, so its stroke translates directly into deflection; and it carries a high-pressure filter at its inlet — per AMM, the LRUs include a high pressure filter, a biased servovalve — protecting the precision servovalve from contamination in the hydraulic fluid.

7.1 The biased zero — why "no command" is not "centre"

The behaviour of every spoiler failure hinges on one design choice in the servovalve: its rest position is not neutral but biased toward retract. Per AMM 27-64-00:

The servovalve has a biased zero. When the spoiler servo control is not activated the input signal is equivalent to the electrical zero, its spool valve is open at 20 percent of its maximum travel. In this position it lets hydraulic fluid pressure into the small chamber of the servoactuator to hold the unit in the retracted position.

So with no command and live hydraulics, the valve sits 20% open toward retract and holds the panel down. The normal baseline that makes this work is worth fixing too. Per AMM 27-64-00:

Hydraulic pressure is applied to the bypass valve and to the plunger which holds the blocking valve in the open position. Thus the two actuator chambers are connected to the servovalve control lines.

Hydraulic pressure itself holds the blocking valve open so the servovalve can command the chambers. The instant that pressure is lost, the blocking valve closes under spring force and isolates the actuator — which is exactly why a hydraulic loss freezes the panel rather than letting it drive.

7.2 Electrical loss → driven to retract

Per AMM 27-64-00:

If there is a failure of the electrical power to the spoilers the servovalve input signal is 'nulled'. The servovalve is biased towards the retraction mode and the spoilers are retracted.

FCOM states the same outcome at system level. Per FCOM DSC-27-10-20:

The system automatically retracts the spoilers to their zero position, if it detects a fault or loses electrical control.

Electrical loss = actively retracted. Hydraulics are still present, so the biased valve drives the panel down to zero. This is the safe side: a panel that loses its signal cannot stand up unbidden and cause unwanted lift loss or roll.

7.3 Hydraulic loss → frozen, then bled down by airflow

Per FCOM DSC-27-10-20:

If the system loses hydraulic pressure, the spoiler retains the deflection it had at the time of the loss, or a lesser deflection if aerodynamic forces push it down.

The AMM gives the valve sequence that produces this "freeze". Per AMM 27-64-00:

When the hydraulic pressure decreases: ‐ the check valve closes and isolates the servovalve from the hydraulic pressure, ‐ the plunger moves and the blocking valve closes, thus the actuator piston cannot extend, ‐ the bypass valve moves rearward under spring pressure, this isolates the piston chambers from the servovalve control lines, ‐ both chambers of the actuator piston are connected to let the servo actuator retract. ‐ aerodynamic forces retract the spoiler. The anti-cavitation valve and the calibrated orifice in the bypass valve stop a fluid underpressure on the extension side of the actuator piston.

With pressure gone the check valve seals the servovalve off, the blocking valve locks the piston against further extension, and the bypass valve cross-connects the two chambers so the panel can only be pushed down by the airflow — passively, and to a lesser deflection. The anti-cavitation valve and the calibrated orifice keep air out of the small chamber if the return line ruptures. The panel cannot be actively positioned at all.

[!warning]- A lost hydraulic system does not instantly fold its spoilers away.

The instinct "lose Yellow, panels 4 and 6 snap flat" is wrong. Electrical loss retracts a panel; hydraulic loss freezes it at whatever deflection it held, after which the airflow eases it down. So after a Yellow loss in the cruise, panels 4 and 6 sit where they were and trail down gradually — they are not actively commanded to zero. Keep the two failure directions separate; they are genuinely opposite.

7.4 The 4-and-6 exception in roll

The roll function carries one exception the speedbrake function does not. Per FCOM DSC-27-10-20:

When a spoiler surface on one wing fails, the symmetric one on the other wing is inhibited (except for spoilers 4 and 6).

For roll, a failed panel normally disables its symmetric partner to avoid an asymmetric rolling moment — except panels 4 and 6. Panels 4–6 are the MLA panels (§2), so disabling a 4 or 6 with its opposite number would weaken the symmetric load-alleviation capability. The design chooses to protect MLA symmetry over absolute roll symmetry on those two panels; the load-alleviation mechanism is developed in Load Alleviation.

8. Retracting the speedbrake — two flying techniques

The hardware above is how the panel moves; the FCTM adds how the pilot should move it. Retracting the speedbrake is the same lever action, but it carries a trap in two situations.

8.1 Gentle retraction in Direct Law — the pitch coupling

Raising or lowering the speedbrake changes the wing's lift distribution, so it is inherently a pitch disturbance source. In Normal Law the fly-by-wire auto-trim absorbs that coupling and the pilot feels nothing. In Direct Law the auto-trim is gone, the coupling is exposed, and an aft centre of gravity makes the aircraft more sensitive still. Per FCTM AOP-10-30-20:

DIRECT LAW — The PF must avoid performing large thrust changes, or sudden speedbrake movements, particularly if the center of gravity is aft. If the speedbrakes are out, and the aircraft has been re-trimmed, the PF must gently retract the speedbrakes to give the aircraft time to re-trim, and thereby avoid a large nose down trim change.

With the speedbrake out you have trimmed against its pitch effect; snap it in and the lift returns abruptly while the THS is still trimmed for "speedbrake out" — the nose drops sharply, more so with an aft CG. So in Direct Law retract the speedbrake slowly and let the trim catch up, treating it as a pitch-disturbing surface, not a deceleration knob. The path into Direct Law and manual trim is covered in Direct Law.

8.2 Regaining the descent profile — don't make the autothrust fight the boards

There is a normal-operations cue as well. When you are above the FMS descent profile and using the speedbrake to increase descent rate, retract it as the profile is regained — otherwise the autothrust ends up adding thrust against the still-raised boards, an energy stand-off that wastes both fuel and drag. Per FCTM PR-NP-SOP-170:

When regaining the descent profile, the speedbrakes should be retracted to prevent the A/THR applying thrust against speedbrakes. If the speedbrakes are not retracted, the "SPD BRK" message on the ECAM memo becomes amber and "RETRACT SPEEDBRAKES" is displayed in white on the PFD.

Speedbrake and thrust are opposite energy levers — one adds drag, the other adds thrust. Leave the boards out once the profile is regained and the system nags you with an amber SPD BRK memo and a RETRACT SPEEDBRAKES prompt on the PFD. The full reading of the amber SPD BRK memo and other ECAM/PFD cues belongs to Controls and Indications.

9. Operational view and failures

Six scenes fly the architecture:

1. Top of speed, need to descend — pull the speedbrake lever; with no inhibition active the PRIM commands all six panels symmetrically up (≤30°); drag rises, descent rate increases; the F/CTL page shows six panels raised together.
2. Bank to correct heading during the descent — with the speedbrake out you roll; on panels 2–6 the roll and speedbrake orders sum past the in-flight limit, so the system lowers the opposite wing's speedbrake (§5). The bank builds normally; the deceleration briefly eases.
3. Above the descent profile, then regaining it — retract the speedbrake as the profile is regained, or the autothrust pushes against the boards and you get an amber SPD BRK + RETRACT SPEEDBRAKES cue (§8.2). If you have degraded into Direct Law, retract gently to avoid a nose-down trim change, especially with an aft CG (§8.1).
4. Approach, high AoA triggers low-speed stability or AoA protection — even with the lever pulled, the system auto-retracts the speedbrake to protect lift margin (§6). To use it again you must wait for the condition to clear, reset the lever, and wait 5 s.
5. Go-around, thrust levers to TOGA — a lever above MCT inhibits and retracts the speedbrake (§6), so it cannot fight the climb thrust.
6. Cruise, one hydraulic system lost (e.g. Yellow) — panels 4 and 6 freeze and trail down under airflow, but the four panels on Green (1, 5) and Blue (2, 3) remain. Each system feeds exactly two panels per wing, so roll and speedbrake authority are reduced but never lost; the aircraft remains fully controllable for landing. The remaining-surface picture for combined hydraulic failures is in EFCS Computer Failures and Degradation.

Indication and dispatch. A failed panel (or pair) presents as F/CTL SPLR FAULT. Dispatch with spoilers inoperative is governed by the operator MEL (item 27-64-01): the panels must be inoperative as a symmetric pair, locked in the retracted position, and a flight-manual performance-loss procedure must be applied — and some operators add airport restrictions (notably high-altitude aerodromes), reflecting the reduced lift-dump and deceleration capability. Treat the specific dispatch conditions as operator-specific. Note that the ground-spoiler fault (F/CTL GND SPLR FAULT) is a separate ECAM with its own dispatch path, covered in Ground Spoilers and Function Allocation.

Self-test

[!note]- Q1. There are six spoilers per wing — why does spoiler 1 not assist roll, and how does it differ from panels 2–6 in deflection and travel?

Roll is performed by panels 2–6 (FCOM: "five spoilers"); spoiler 1 serves only the speedbrake and ground-spoiler functions, never roll and never MLA. Because it is a dedicated deceleration panel it has the shortest stroke (50.7 mm) and the smallest speedbrake limit (25°), versus 69.5 mm / 30° for panels 2–6. The speedbrake uses all six for maximum drag area; roll uses only 2–6 because roll is a single-sided demand spoiler 1 is not wired for.

[!note]- Q2. How are the six panels split across the hydraulic systems and computers, and why does losing one hydraulic system cost only two panels per wing?

Hydraulics are paired: Green 1 & 5, Blue 2 & 3, Yellow 4 & 6. Computers are interleaved panel by panel: 1 & 2 → PRIM 3, 3 → SEC 2, 4 → PRIM 2, 5 → PRIM 1, 6 → SEC 1. Since each hydraulic system feeds exactly two panels per wing, losing any one system removes only those two — four panels remain — so roll and speedbrake authority degrade but are not lost. Spoilers are also the one surface not taken over by another computer, so a fault means those panels stay unavailable.

[!note]- Q3. With both a roll order and a speedbrake order on panels 2–6 exceeding the in-flight limit, what does the system sacrifice, and why that choice?

Roll has priority. Rather than trimming the saturated panel, the system retracts the symmetric speedbrake on the opposite wing until the difference between the two wings equals the roll order. Roll comes from the difference between the wings, so lowering the opposite wing restores the needed difference without asking the maxed-out panel for more. The cost: while you hold bank, the speedbrake's deceleration is temporarily reduced.

[!note]- Q4. After an electrical failure and after a hydraulic failure, where does a spoiler end up — and why are the two opposite?

Electrical loss → actively retracted to zero: the servovalve's biased zero (spool 20% open toward retract) drives the panel down while hydraulics are still available — the safe side. Hydraulic loss → frozen at the loss-instant deflection, then bled down by airflow: the check valve isolates the servovalve, the blocking valve locks the piston, the bypass valve cross-links the chambers, so the panel has no active authority and only the airflow eases it down. The opposite outcomes both stem from the biased valve — it can only drive the panel when both power types are present.

[!note]- Q5. In which five cases is speedbrake extension inhibited, and why does re-extending need a lever reset plus a 5-second wait?

Inhibited if: MLA active, AoA protection active, low-speed stability active, at least one thrust lever above MCT, or alpha floor active. The three low-speed/high-AoA cases protect lift margin; MLA has commandeered panels 4–6; thrust above MCT contradicts decelerating. If inhibited while extended, the panels auto-retract and stay retracted until the condition clears — and they will not re-extend until the crew resets the lever and 5 s have passed, a deliberate re-arming step so the boards cannot pop out unnoticed.

[!note]- Q6. Why must the speedbrake be retracted gently in Direct Law, and what cue appears if you forget to retract it after regaining the descent profile?

In Direct Law there is no auto-trim, so the speedbrake's pitch coupling is exposed; a sudden retraction returns lift abruptly while the THS is still trimmed for "boards out", producing a sharp nose-down trim change — worse with an aft CG. Retract slowly and let the trim catch up. After regaining the descent profile, an unretracted speedbrake makes the autothrust add thrust against the boards; the system shows an amber SPD BRK ECAM memo and RETRACT SPEEDBRAKES on the PFD.

Key takeaways

References

Per FCOM DSC-27-10-20 (Roll control — two ailerons and five spoilers per wing, 35° maximum spoiler deflection; roll allocation SEC 3 & 6 / PRIM 2, 4 & 5; spoiler actuation — auto-retract to zero on fault or electrical loss, retain deflection on hydraulic loss, symmetric-failure inhibit except panels 4 and 6. Speedbrake and ground spoiler control — speedbrake lever, speedbrakes are spoilers 1 to 6; five inhibition conditions with auto-retract / lever reset / 5 s re-arm; symmetric speedbrake-failure inhibit; maximum deflection 25° spoiler 1 / 30° spoilers 2–6, reduced in CONF 2/3/FULL; roll priority over speedbrake). Per AMM 27-60-00 (six spoilers per wing and their three functions; speedbrake signal chain lever → FCPC → FCSC → servocontrols; one FCPC or one FCSC controls each servocontrol). Per AMM 27-64-00 (electrohydraulic servocontrol per spoiler; LVDT position feedback and fault detection; hydraulic allocation Green 1 & 5 / Blue 2 & 3 / Yellow 4 & 6; controlled travel 50.7 mm spoiler 1 / 69.5 mm spoilers 2–6; double-acting servo, two eye ends, high-pressure filter and biased servovalve; normal operation with blocking valve held open; biased zero at 20% spool travel; electrical-failure retraction; hydraulic-failure valve sequence — check valve, blocking valve, bypass valve, anti-cavitation valve and calibrated orifice). Per ASM 27-98-31 thru 35 (per-panel control-computer assignment: spoilers 1 & 2 / FCPC 3, 3 / FCSC 2, 4 / FCPC 2, 5 / FCPC 1, 6 / FCSC 1). Per FCTM AOP-10-30-20 (Direct Law — gentle speedbrake retraction to avoid a large nose-down trim change, aft-CG sensitivity). Per FCTM PR-NP-SOP-170 (regaining the descent profile — retract speedbrakes to prevent A/THR applying thrust against them; amber SPD BRK memo and RETRACT SPEEDBRAKES PFD prompt). Hydraulic-system dependency per ATA-29. Dispatch is governed by the operator MEL (item 27-64-01) and is operator-specific. The asymmetric rise-one-wing roll behaviour and the MLA-symmetry rationale for the 4-and-6 roll exception are integrative synthesis consistent with the FCOM allocation and the AMM function list, not single verbatim manual statements.

Independent study material, not an Airbus publication and not endorsed by the manufacturer. Always defer to the current operator FCOM, FCTM, and QRH for operational use.