Bryant Furnace Ignition Codes (14, 34): Meaning & Fixes

Two Bryant codes own the whole ignition story: 34, ignition proving failure, when a single attempt cannot light the burners and prove flame; and 14, ignition lockout, when four attempts in a row have failed and the control shuts gas heating down for 3 hours. This article walks the Bryant ignition sequence step by step, gives each code its own section with Bryant's documented causes and checks, and is plain about the boundary Bryant itself draws: the sequence can be watched and the simple fuel facts confirmed, but gas pressure, gas valves, igniters and flame-sense measurements are licensed-technician work.

The Bryant ignition sequence, step by step

Every Bryant ignition code is a report about one step of a fixed sequence, and the variable-speed 355MAV guide times it out in full. The wall thermostat calls for heat, closing the R to W circuit. The control first performs a self-check and verifies that the low-heat and high-heat pressure switch contacts are open — if one is already made, the sequence stops at code 23 before anything turns. The inducer motor then starts and slowly increases speed; when the low-heat pressure switch closes, the control notes the inducer RPM — Bryant uses that number to judge the vent system's resistance and set the inducer speed for the cycle — and a 25-second prepurge begins, clearing the heat exchanger of any leftover gas. (If the cycle starts on high-heat, the control keeps increasing inducer speed after the low switch closes until the high-heat switch closes, and the 25-second prepurge runs from there.) At the end of prepurge comes the hot-surface igniter's warm-up: 17 seconds, in Bryant's timing, for the igniter to reach lighting temperature. Then the trial for ignition: the gas valve opens, gas flows to the burners and is ignited; after 5 seconds the igniter is de-energised and a 2-second flame-proving period begins, during which the flame-proving sensor electrode must report flame to the control. Prove flame, and the control holds the gas valve open and starts the blower-on delay — 60 seconds to low-heat airflow on the 355MAV (30 seconds on the 355BAV), 35 seconds to high-heat airflow — while the inducer settles to its running speed. Fail to prove flame inside that 2-second window, and the control closes the gas valve and repeats the whole ignition sequence, up to three more trials for ignition, before going to ignition lockout. Single-stage Bryant models run the same order — prove the vent, purge, warm the igniter, open the valve, prove the flame — with the timings on their own door labels. When you watch a failing furnace, name the step it dies at: inducer never runs, igniter never glows, burners never light, or burners light and die in seconds. Each death is a different branch of the codes below.

Code 34: ignition proving failure

Bryant's definition is exact: if flame is not sensed during the trial-for-ignition period, the control repeats the ignition sequence 3 more times before lockout 14 occurs — so a flashing 34 is an attempt in progress or just failed, not yet a lockout. There is a second 34 pattern owners misread: if the flame signal is lost during the blower on-delay period (flame proved, burners running, then the signal drops as the blower is about to start), the blower will come on for the selected blower off-delay — a furnace that lights, runs briefly, blows air and stops is telling that version of the story. Bryant's service-label instruction for 34 is 'check the following items first before proceeding to the next step', and the list is deliberately mundane at the top: the gas valve turned off, and the manual shut-off valve closed. Below those come the flame-sense facts Bryant quantifies: oxide buildup on the flame sensor (the documented cleaning is fine steel wool), and proper flame sense current — 0.5 microamps DC minimum, 4.0 to 6.0 nominal (the 330JAV label specifies the nominal in high heat). Then the ignition hardware and supplies: a defective hot-surface igniter, a defective gas valve, low inlet gas pressure, inadequate flame carryover or rough ignition (flame lighting at the first burner and failing to spread cleanly across the rest). And finally the two grounding rules Bryant prints in capitals-adjacent emphasis: the green/yellow wire MUST be connected to furnace sheet metal, control ground continuity matters, and the flame sensor must NOT be grounded — flame rectification is a tiny electrical signal, and a bad ground reference makes a healthy flame invisible to the control. Treat any single 34 as evidence: note which step failed and whether the burners lit at all before you touch anything.

Code 14: ignition lockout

Code 14 is 34's arithmetic completed: the 340AAV label states it directly — 'System failed to ignite gas and prove flame in 4 attempts. Control will auto-reset after 3 hours. See status code 34.' Four attempts means the initial trial plus the three repeats code 34 promises; when the fourth attempt also fails to prove flame, the control stops offering gas to a furnace that cannot light it, and gas heating is locked out. The auto-reset after 3 hours is the control giving the world one fresh chance — a gas supply restored, a sticky condition cleared — not a diagnosis and not a repair. Diagnostically, 14 adds nothing to 34's cause list, and Bryant structures the manuals that way on purpose: 14's entire instruction is to refer back to 34. What 14 does add is a decision. A furnace that has burned four ignition attempts in a row has run a controlled experiment and reported the result: something in the chain — fuel, igniter, flame sense, ground, carryover — is measurably wrong. The owner's remaining move is exactly one clean retest after the genuinely simple facts are confirmed (gas valve on, manual shut-off open, other gas appliances working); if that retest also walks to lockout, the furnace has answered twice, and further resets only repeat the experiment at the heat exchanger's and the igniter's expense. On an Evolution system, also read the event at the Connex wall control and the stored history at the board before the lockout clears: a season of quiet 34s graduating to 14s is a degrading igniter or sensor announcing itself, and the history is where that slope is visible.

The causes both codes share — Bryant's check-first list as a system

Because 14 is simply 34 four times, the two codes share one cause system, and it helps to see Bryant's list grouped the way a technician walks it. Fuel delivery first: the gas valve's own switch left off after other work, the manual shut-off valve closed, a defective gas valve that receives its signal and does not open, and low inlet gas pressure — the furnace cannot light gas it does not have, and Bryant's sequence has no way to distinguish 'no gas' from 'no ignition' until someone checks the supply. Ignition source second: the hot-surface igniter, which Bryant's door-label flow isolates with the component test — run the test and watch whether the igniter glows orange/white by the end of its 15-second warm-up; a dark igniter with voltage present at its harness is a replaced igniter, and a lit igniter that never produces flame points back at fuel. Flame proof third, and the subtlest: the flame sensor (the flame-proving sensor electrode), whose oxide coating Bryant cleans with fine steel wool; its signal, measured in microamps DC against Bryant's 0.5 minimum and 4.0–6.0 nominal; the green/yellow ground wire that MUST be connected to the furnace sheet metal; control ground continuity; and the sensor's own insulation — a flame sensor that is grounded cannot report. A burner can light perfectly and still generate a 34 when this third group is wrong, which is why 'the burners light for three seconds and quit' is the flame-sense branch until proven otherwise. Carryover last: flame that lights at the first burner and fails to spread — inadequate carryover or rough ignition on Bryant's list — produces ragged 34s and belongs to burner and manifold-pressure work. Note what is NOT on Bryant's 34 list: the control board itself appears only at the end of the diagnostic flow, after gas, igniter and sensor have been eliminated one by one.

Isolating the igniter: Bryant's component test

One diagnostic door on a Bryant furnace is built for exactly the 34/14 question 'is it the igniter?', and Bryant prints it on the door label as the component test. The idea is to remove the thermostat and the gas from the experiment: with the thermostat call removed (the labels direct removing the R thermostat connection), the control is put into its test sequence — on the variable-speed models via setup switch SW1-6, on the 987M by removing the R wire, setting SW1-6 ON and holding the blower door switch closed. The control then runs its components in a fixed order without attempting to heat: the inducer motor comes on and stays on, the hot-surface igniter is energised for about 15 seconds, and the blower motor runs for about 15 seconds, after which result codes flash (11, 25, 41, 42 on the 355MAV; the 987M's list adds 35). For the igniter question, the 340AAV label's flow is satisfyingly direct: run the component test and watch the igniter — does it glow orange/white by the end of the 15-second warm-up period? If it does not, the flow unplugs the igniter harness and repeats the test while checking for 115 volts between the igniter pin and neutral on the control: voltage present for the warm-up period with a dark igniter condemns the igniter; no voltage sends the diagnosis to the control and its connections. A glowing igniter, meanwhile, acquits the ignition source entirely and moves a 34 investigation back to fuel delivery and flame sense — the two branches the test cannot see, because it never opens the gas valve. That boundary is worth respecting in both directions: the component test is a technician's procedure on most Bryant models (it involves the door switch held closed on a live furnace, which Bryant wraps in electrical-shock warnings), but knowing it exists tells you what a competent Bryant service visit will do in its first ten minutes — and why a quote that replaces the igniter, the gas valve and the control in one visit, with no test in between, has skipped Bryant's own order of operations.

Reset and lockout behaviour: what resets, what waits, what must never be forced

Bryant gives ignition failures three exits, and the manuals document all of them. The automatic exit: lockout 14 clears itself after 3 hours, and the furnace will try again on the next call for heat. The manual exit: the 355MAV sequence of operation states that the ignition lockout 'will be reset automatically after three hours, by momentarily interrupting 115 vac power to the furnace, or by interrupting 24 vac power at SEC1 or SEC2 to the furnace control' — a power cycle is a legitimate Bryant reset for this lockout, once the cause has been addressed. The third exit is the one that does not exist: there is no way to talk the control out of its four-attempt arithmetic, and there is no component to jumper — the pressure switches, limit circuit and flame proof are wired into the sequence precisely so that gas is never offered without its proofs. Use the reset budget the way Bryant intends. One power-cycle retest, after the fuel facts are confirmed, is a diagnostic step: a furnace that lights and holds had a transient (air in a reopened gas line is the classic); a furnace that walks back through 34 to 14 has replicated its fault under controlled conditions, and that replication is the finding. Repeatedly cycling power to harvest a few minutes of flame from a furnace that cannot prove it is the behaviour the lockout was designed to prevent — every attempt opens a gas valve on an unproven burner. And if the ignition failures ever arrive with company — code 22 (flame proved while the valve should be off) or code 21 (gas heating lockout) in the history — stop entirely: those are gas-valve integrity reports, and they outrank everything else on this page.

When ignition work becomes a licensed technician's job

Bryant's own literature draws this line in the same place this page does. The troubleshooting flow behind code 34 asks its questions in an order that moves steadily into the gas train: Is 24 volts present at the gas valve? Does the valve open and allow gas to flow? Do the main burners ignite, and do they stay on? What is the flame sensor current during the trial for ignition? Answering those questions means working at an open burner compartment with a meter on a live gas valve and a manometer on the manifold — the 355MAV guide even notes, in its low-heat no-ignition branch, adjusting and setting low-heat manifold pressure and re-orificing if necessary. That is licensed gas work, and it begins sooner than most owners expect: effectively everything past 'the gas is on and the burners do or do not light' belongs to a technician. Call one when: the furnace reaches lockout 14 at all after the simple fuel checks; you smell gas at any point (stop first, call the gas utility as well); the igniter does not glow in the component test; the burners light but will not stay lit (the flame-sense branch — sensor cleaning itself is listed in Bryant's flow, but the microamp measurement that proves it is meter work); or the stored history shows 21 or 22 alongside the ignition codes. What you can hand the technician that genuinely shortens the visit: the code (34 vs 14), which step the sequence died at, whether the igniter glowed, how far the burners got, and the history pattern. On Bryant ignition faults, that observation set is most of the diagnosis — the rest is instruments.

Key takeaways

• Count the attempts: a furnace showing 34 is still trying (up to 4 attempts total); one showing 14 has finished trying and is in its 3-hour lockout. The count tells you which conversation you are having.
• Watch one full attempt before resetting anything, and classify it: no igniter glow, glow but no flame, flame for seconds then out, or only the first burner lighting. That single observation routes the whole Bryant check list.
• Photograph the flashing code before any power cycle — on older Bryant boards, interrupting power erases the stored fault history along with the lockout.

Resetting the Bryant furnace

Bryant's lockouts run on documented timers, and they are not interchangeable. Limit lockout (13) and ignition lockout (14) auto-reset after 3 hours. Blower motor lockout (15, variable-speed boards) also auto-resets after 3 hours. Control circuitry lockout (45) auto-resets after 1 hour, and Bryant adds: reset power to clear the lockout — if 45 repeats, replace the control. Gas heating lockout (21) is the exception with NO auto-reset: Bryant's instruction is to turn off power and wait 5 minutes to retry. The pressure-switch behaviour under 31/32 (inducer shuts off for 15 minutes after the switch has been open 5 minutes, then retries) is a retry pause, not a lockout. Everything else on the Bryant chart is a condition code that clears when its condition clears: 33 ends when the limit or rollout switch remakes as the furnace cools (the rollout switch itself is manual-reset hardware — no timer resets it, and Bryant requires its cause to be corrected before it is reset), and 24 ends only when the 24-volt short is repaired and the fuse replaced. Interrupting power restarts a Bryant control, but on the older boards it also erases the stored fault history — read the history first, and never treat a cleared timer as a repaired furnace: an unfixed cause simply walks the control back into the same lockout.

Frequently asked questions

Related guides

Facts on this page were checked against manufacturer documentation: shoredist.com · manuzoid.com · www.manualshelf.com · www.manualsdir.com · www.manualshelf.com · manualsdump.com · manualsdump.com · www.manualsdir.com · www.manualslib.com · manuals.plus