The Five Quiet Computers: Why SECs and FACs Finish A320 Oral Exams

The ELACs take the architectural spotlight. The questions that finish orals come from the other five.

By Paul Toensing  |  Custom GPT Solutions  |  June 2026  |  ~12 min read

Type rating candidates spend most of their flight-control study time on the ELACs. The protection envelope, normal law, the sidestick authority pilots feel every flight - it all lives in those two computers, and the attention is, frankly, earned. But the questions that finish A320 oral exams are not about the ELACs. They are about the three SECs and the two FACs, the five quiet computers in the architectural corner that do most of the unglamorous work while the spotlight stays elsewhere. An examiner who has watched a few hundred candidates can tell you exactly where the orals tend to come unstuck, and it is rarely on the famous systems. The candidate who walked in confident about pitch and roll normal law walks out wondering how the conversation ended up on spoiler-to-SEC mapping.

This article walks the SEC and FAC architecture at the level an examiner expects: what each system does, what the fixed mappings actually are, and what specifically goes wrong in the two cascades that catch candidates who studied surface-level rather than depth-level. It is a longer-form complement to the FAQ-structured AEO page at /a320-secs-facs-deep-dive/, which answers the same territory in search-intent question blocks. For the foundational seven-computer overview, start with /a320-elac-sec-fac-foundation/. For the ELAC architecture in depth, see /a320-elacs-deep-dive/.

The SECs Run the Spoilers, and the Mapping is Not Intuitive

The three SECs (Spoiler Elevator Computers) control specific spoiler panels on each wing. The assignment is fixed in the FCOM, and the architecture pays no attention to what would be convenient to memorize. SEC 3 controls the No. 2 spoilers. SEC 1 controls the No. 3 and No. 4 spoilers. SEC 2 controls the No. 5 spoilers. The numbering for the SECs does not line up with the numbering for the spoiler panels, the mapping is not symmetrical, and it is not in ascending order. Somewhere in Toulouse, a design engineer made these choices, and the rest of us live with them. If a candidate's mental model has SEC 1 controlling spoiler 1, the mental model is wrong, and an examiner will know it within ten seconds.

SEC 1 and SEC 2 carry the backup pitch path. SEC 3 controls spoiler panel 2 only and does not participate.

The distribution matters because it means no single SEC failure produces a catastrophic loss of spoiler authority. Each SEC failure removes only the panels assigned to that computer, leaving meaningful roll and lift-dumping capability through the survivors. The system is more elegant than the numbering convention suggests, which is something of an Airbus signature.

Beyond the spoilers, SEC 1 and SEC 2 carry partial elevator authority as the backup pitch path if both ELACs fail. If neither ELAC is available, pitch control shifts to whichever SEC the circuit logic supports at that moment. SEC 3 does not participate in the pitch backup - it controls only its assigned spoiler panel. The SECs also participate in ground spoiler extension, which uses the published ground, lever, and thrust conditions from the FCOM. The popular candidate story that a single SEC failure at touchdown collapses autobrake through loss of automatic lift-dumping is a tidy mental model. The FCOM disagrees with it. AUTO BRK in the abnormal status is affected only if at least 2 SECs fail, not 1. The architecture has more layers than the candidate's invented causation usually allows for, and the examiner who hears the invented version asks where the candidate read it.

What does not belong to the SECs: the ailerons (ELAC territory in normal law), the rudder (mechanical at the pedals, with the FACs providing the electrical augmentation functions on top), and the flight envelope protections (ELAC- and FAC-computed). The SECs are the surfaces-on-the-wings computers plus the pitch backup if the ELACs check out.

Source: FCOM DSC-27-10-20, p.10-11/16; PRO-ABN-F_CTL, p.62/92.

The FACs and the Two-Path Rudder

The two-path rudder model. Mechanical pilot input through the pedals on one path. FAC augmentation - yaw damping, turn coordination, rudder trim, travel limitation - on the other.

The two FACs (Flight Augmentation Computers) are not symmetric to the SECs. They do entirely different work, and the first thing to get right about them is the two-path rudder control model. The pilot's basic rudder control path is mechanical through the pedals. The linkage from the feet to the rudder surface does not pass through any computer, and the rudder is hydraulically actuated by three servojacks. But the FACs do control the rudder's augmentation functions: yaw damping, turn coordination, rudder trim, and rudder travel limitation. Two paths, both real, both required.

The phrase "FACs do not control the rudder" is a common candidate shortcut that loses points in an oral, because the examiner is listening for the candidate who understands what the FACs do contribute, not the candidate who memorized a half-truth. The right framing is mechanical pilot input on one path and computer augmentation on the other. The two paths coexist by design, and the design exists because the engineer who wired the rudder did not want a single computer failure to disable yaw control entirely. The pilot keeps the pedals.

FAC work falls into two broad categories. The first is electrical rudder control: yaw damping, turn coordination, rudder trim, and rudder travel limitation (the variable mechanical limit that reduces maximum rudder deflection at high speed). The pilot retains direct mechanical pedal authority to the rudder surface, and the FACs add the stabilization layer on top of that. The result is a rudder that behaves the way pilots expect it to behave, which is exactly the kind of helpful invisible work that pilots stop noticing until it stops working.

The second category is flight envelope computation. The FACs compute the dynamic characteristic speeds you read on the PFD speed scale (VLS, VFE, VMO, VMAX, F-speed, S-speed, green dot), the alpha-floor command logic, the windshear detection and warning, and the low-energy "SPEED SPEED SPEED" warning. The air data itself - raw airspeed, altitude, angle of attack, temperature - comes from the ADIRUs. The FACs are the processing layer that takes the raw ADIRU data and combines it with weight and configuration information to compute the speeds and warnings the pilot actually sees.

This distinction matters because it changes the right answer to one of the more common examiner traps. "If both FACs fail, can we still get airspeed?" Yes. Raw airspeed comes from the ADIRUs and remains on the PFD. It is the characteristic speed overlays that disappear, not the airspeed itself. The candidate who panics and answers "no, we lose airspeed" has just told the examiner they do not know where the air data lives, which is a problem because the examiner is about to ask the follow-up question for sport.

Source: FCOM DSC-22_40-10, p.1/2; DSC-27-10-20, p.13-14/16.

Single SEC Failure - The SEC 1 Speedbrake Restriction

An individual SEC failure triggers an amber ECAM caution and a panel fault light. The flight control laws remain intact. The airplane stays in normal law with full envelope protections active. The specific spoiler panels assigned to the failed computer are lost or retracted, but the roll authority through the remaining panels and the ailerons is preserved.

The case that deserves specific attention is SEC 1 failure. SEC 1 controls the No. 3 and No. 4 spoilers, which are the primary speedbrake surfaces. When SEC 1 is affected, the ECAM presents the published restriction SPD BRK (IF SEC 1 AFFECTED) DO NOT USE. This is not a polite suggestion.

The FCOM's stated reason for the restriction is more interesting than most candidates' invented one. The manual says: VLS would not be corrected if speed brakes 2 extend, because no speed brake position is sent to the FACs. In other words, the FACs do not know the speedbrakes are out, so they cannot adjust VLS to account for the higher drag and lower lift coefficient that speedbrake extension would normally produce. The candidate who invents a story about symmetric deployment or asymmetric drag has the wrong mental model, and the examiner will press on it because the manual reason is right there and easier to defend. Quote the FCOM. Stop. The candidate who quotes a homemade systems explanation in an oral is asking the examiner to find a hole in it, and on a busy day the examiner will.

For SEC 1 or SEC 2 failure specifically, the elevator backup capability is also degraded. One of the two SECs that can provide pitch backup is gone, and only the remaining one is available if both ELACs subsequently fail. The pilot does not feel this in normal operation because the ELACs are still flying the airplane. But the layered architecture has lost a layer, and the right answer in an oral names the layer that is gone.

Source: FCOM PRO-ABN-F_CTL, p.59/92.

Triple-SEC Failure - Two Axes, Two Different Laws

The triple-SEC failure is one of the two cascades where the quiet ones become loud. F/CTL SEC 1+2+3 FAULT is where homemade mental models go to die in oral exams, because the systems text is more specific than candidate intuition usually allows for. The FCOM wording is worth quoting carefully, and the candidates who quote it cleanly tend to do better in the next twenty minutes of the oral.

Pitch alternate law and roll direct law become active simultaneously. Two axes, two different laws

On 3 SEC failures, the more exact systems statement is: pitch alternate law and roll direct law become active. Two axes, two different laws. The ECAM status later shows ALTN LAW: PROT LOST, with direct-law reversion triggered by the published landing and configuration logic. The ECAM status is the shorthand candidates memorize. The axis-specific systems text is what earns points in an oral. The candidate who says "alternate law" answers the question. The candidate who says "pitch alternate, roll direct" answers it the way the FCOM does, and the examiner notices.

The direct law reversion that follows is governed by landing configuration logic, not by autopilot state alone. With the autopilot engaged, the trigger for direct law is WHEN L/G DN AND AP OFF - both landing gear extension and autopilot disengagement are required. With the autopilot not engaged at the time of failure, the trigger is WHEN L/G DN - landing gear extension alone produces the reversion to direct law. There is also a configuration substitution worth knowing: ELAC uses slats/flaps CONF 2 as the trigger instead of landing gear extension when the substitution logic is active.

The common wrong answer in oral exams is "direct law on AP disconnect." That is incorrect. The trigger is tied to landing configuration, not autopilot state alone. A candidate who tells an examiner "the autopilot is acting as a temporary law stabilization buffer" is going to get a polite correction, followed by a more rigorous follow-up sequence, and the rest of the oral starts to feel longer than it is.

Bad mental models in oral exams compound. The first wrong answer becomes the foundation for a sequence of wrong follow-ups, and the examiner gets to choose how deep into the swamp the candidate goes. The triple-SEC trigger is the kind of detail worth studying with precision, because the candidate who quotes the FCOM has nothing to defend, and the candidate who improvises is defending invented architecture against a person paid to know the real one.

Source: FCOM PRO-ABN-F_CTL, p.59-60/92.

Single FAC Failure - Less Invisible Than It Looks

A single FAC failure (AUTO FLT FAC 1(2) FAULT) is handled gracefully by the surviving FAC, but it is not operationally invisible the way pilots sometimes assume. All FAC functions transfer to the surviving computer. Yaw damping, turn coordination, rudder trim, rudder travel limitation, characteristic speed computation, alpha-floor command, windshear detection, and low-energy warning all continue. The pilot does not feel a handling change. The airplane flies normally. So far, so good.

But the status message tells the pilot what just changed. BOTH PFD ON SAME FAC. CAT 3 SINGLE ONLY. The redundancy that allowed autoland to CAT 3 DUAL is gone. Any approach to landing minima below CAT 3 SINGLE requires the redundancy that no longer exists. The right oral exam answer to "is single FAC failure invisible?" is: "Handling is largely unchanged because the surviving FAC picks up everything, but operationally we have lost autoland redundancy and we now have both PFDs reading from the same FAC." That is the answer that demonstrates the candidate understands operational consequences beyond control feel. The candidate who says "basically nothing" answers the question. The candidate who names what changed answers the question well.

Dual FAC Failure - Effectivity Matters, FCOM Wins

Six simultaneous consequences. The ELACs may be healthy but the FAC architecture drives the law reversion regardless.

The dual FAC failure is the second cascade where the quiet ones become loud. It is also where the most candidate-intuitive answer is the wrong one, and where the trainer who simplified is about to be found out. The candidate-intuitive answer is that pitch and roll normal law remain intact on dual FAC failure as long as the ELACs are healthy. The FCOM disagrees. The FCOM wins.

AUTO FLT FAC 1+2 FAULT produces a law outcome that is effectivity-dependent across the A320 fleet. Some MSN standards show F/CTL ALTN LAW (PROT LOST) and may show WHEN L/G DN: DIRECT LAW. Other MSN standards show ALTN LAW indicated with mechanical yaw and no reversion to DIRECT LAW at landing gear extension. The single fleet-wide answer does not exist. The candidate who memorized "dual FAC equals alternate law direct at gear" from a generic training book learns in the oral that the trainer simplified. The candidate who knows which standard the operator's fleet falls into, and quotes the corresponding law table, has the answer that holds up under follow-up questioning.

The FAC 1+2 Fault Trap. The candidate-intuitive answer - healthy ELACs preserve Normal Law - is the wrong one.

What is constant across the effectivity groups is the consequence set. Yaw damping, turn coordination, rudder trim, characteristic speeds on the PFD, alpha-floor (inhibited because alternate law is active), windshear detection, and low-energy warning are all lost. The rudder travel limiter is locked at the value reached at the second failure. The ECAM carries RUD WITH CARE ABV 160 KT, and the FCOM lists the lost functions explicitly: with FAC 1+2 inoperative, the rudder travel limit system, rudder trim control, yaw damper and PFD characteristic speeds are lost. Full rudder travel authority is recovered at slats extension, which is the kind of detail a candidate either knows or invents on the spot.

Speed handling on the approach also depends on the aircraft's effectivity. Some standards show OPERATING SPD COMPUTE on the ECAM status. Others show APPR SPD VREF + 10 KT without that companion line. State the effectivity before stating the procedure. The general principle is to follow what the ECAM gives you and not improvise a computation method that was not published. The candidate who confidently describes a "raw operational data table" technique that does not exist in the operator's FCOM has invented a procedure, and the examiner who recognizes the invention has the next ten minutes of the oral planned. NO DISPATCH applies under MEL.

The handling state is uniquely strange. The airplane is in alternate law with most protections gone. The speed reference the pilot is supposed to fly to depends on which ECAM presentation the aircraft generates. The rudder authority is locked at whatever travel limit was active at the moment of the second failure. Yaw damping is gone, so any rudder input is the pilot's to manage manually. At slats extension on the approach, full rudder travel returns, which sounds like good news until the pilot remembers that yaw damping is still gone and aggressive rudder input on landing rollout in a crosswind is now an unassisted handling problem.

The verified FCOM language for this scenario is the language to quote: RUD WITH CARE ABV 160 KT, and the explicit list of lost functions. Any handling theory beyond that - including natural-aerodynamic-damping arguments about Dutch roll suppression that have made the rounds in trainer notes - is theory, not procedure. The pilot's job in this scenario is to fly by the verified FCOM language, not by what feels intuitively true about rudder dynamics. The discipline in an oral exam is to quote what the manual says and stop there. The discipline in the airplane is the same.

Source: FCOM PRO-ABN-AUTO_FLT, p.8/36, 13/36, 14/36.

Why Alpha-Floor Goes Away, and How Not to Answer That Question Wrong

One of the most common ways candidates lose ground in an oral exam is by trying to explain why alpha-floor is inhibited after dual FAC failure. The temptation is to construct a multi-step causal chain involving autothrust degradation, FAC dependency on the autothrust system, and law state interactions. The FCOM does not invite this much theorizing. It really, really does not.

The published wording is simply that alpha-floor protection is inhibited when alternate or direct law is active. The mechanism is a published rule about which law states permit alpha-floor, not a homemade causal chain. The clean oral exam answer is: "Alpha-floor is inhibited in alternate or direct law per the FCOM, and the dual FAC failure puts the airplane in alternate law, which is why alpha-floor is inhibited." The reasoning ends there. Period. Full stop. Sit down with the satisfaction of a clean answer.

Candidates who improvise a causal explanation are asking to be corrected, because the moment an examiner hears a homemade causal chain, the examiner has a free pull on "so what would happen if [variable] changed?" and the candidate's improvised architecture starts to wobble. The discipline is to quote the rule and stop. Examiners want to see candidates who know the FCOM at the level of its actual wording, not candidates who construct elaborate theories about why the airplane is the way it is. The airplane is the way it is because Airbus designed it that way and wrote it down. The wisest candidate just reads what they wrote.

Why the Quiet Ones Finish Orals

There is an architectural reason candidates underprepare for SEC and FAC questions, and it is not laziness. The ELACs carry the architectural narrative, and study time follows narrative. Normal law lives in the ELACs. The protection envelope lives in the ELACs. The sidestick authority pilots feel every flight comes from the ELACs. The systems that look most important get the most study time, and the systems doing the quiet work get whatever attention is left over, which on the night before a checkride is usually not much.

But examiners know which questions get strong answers and which questions get weak ones. The questions about normal law, pitch architecture, and the protection envelope get strong answers from most candidates because most candidates studied them. The questions about triple-SEC behavior, dual-FAC consequences, the causation chain on alpha-floor loss, the SEC-to-spoiler mapping, and the source of characteristic speeds get weak answers because candidates skimmed past them. The examiner who has the option of asking either type of question, looking at a candidate who has just nailed the famous answers and is starting to relax, has a choice to make. The choice usually does not go in the candidate's favor.

An oral exam can finish on a strong note for the candidate who studied with proportional attention and quotes the FCOM wording, or it can finish on a humbling note for the candidate who improvised. The two scenarios that test the SEC and FAC architecture at depth - triple-SEC failure with its configuration-driven direct law trigger and pitch-alternate/roll-direct axis split, and dual-FAC failure with its effectivity-dependent law table and everything-at-once augmentation loss - are exactly the scenarios that finish orals. The candidates who prepared for them tend to walk out of the room a little more cheerful than the ones who did not. That is also exactly the conversation pilots have in the hotel bar afterwards, and the cheerful ones are usually buying.

Going Deeper

The complete examiner-grade walkthrough of the SEC and FAC architecture, with every FCOM citation, every trap, and every consequence laid out in a conversational format that lets candidates practice thinking out loud before the real oral, lives in the A320 Oral Exam Prepper at customgptsolutions.ai. The airline-specific variants are anchored to each operator's exact FCOM effectivity, which matters specifically for cases like the dual-FAC law table that varies across MSN groups. The Prepper does not pretend to be the FCOM. It asks the questions an examiner asks, in the order an examiner asks them, until thinking on your feet feels less like a guess and more like a habit.

For the question-by-question version of this same territory, see the AEO page at /a320-secs-facs-deep-dive/. For the full Foundation Series, start at /a320-elac-sec-fac-foundation/ and work through to /a320-elacs-deep-dive/.

The original Week 3 article on LinkedIn, with the "Quiet Ones in the Corner" framing, is at

https://www.linkedin.com/pulse/quiet-ones-corner-five-computers-do-unglamorous-work-paul-toensing-9ndoc