Rheem Furnace Pressure Switch Codes (55, 57): Meaning & Fixes

Rheem's pressure switches prove the vent is drafting before gas is offered — and the control proves the switches in both directions, which is why this family has two opposite codes: 55, the switch stuck closed when it should be open, caught before the inducer runs; and 57, the switch stuck open when it should be closed, caught with the inducer running and ignition withheld. This guide gives each code its own section, maps Rheem's wider per-switch family honestly (44/45/46 for the low switch, 55/57 for the high switch, and the modulating models' mid-switch codes left unverified rather than invented), and ends where Rheem's safety logic ends: a pressure switch is never jumpered to get heat.

What the pressure switches prove — and the order Rheem checks them in

Every Rheem heat cycle begins with an interrogation, and the pressure switches are its subject. The switches are air-proving devices plumbed by small hoses to the inducer and vent side of the furnace: when the inducer (the combustion blower) pulls exhaust through the heat exchanger and out the vent, the pressure change closes the switch, and that closure is the control's evidence that the exhaust path is open and drafting. No evidence, no ignition — the gas valve is never offered a cycle whose exhaust has nowhere to go. The interrogation runs in both directions, in a fixed order that explains both codes in this article. First, before the inducer starts, the control verifies the switch is OPEN: a switch already closed at rest is claiming draft exists when nothing is running, which can only mean the switch, its wiring or the pressure around it is lying — that is code 55, and the cycle does not begin. Second, with the inducer running, the control requires the switch to CLOSE within its proving window: a switch still open is reporting that draft was never made or never sensed — that is code 57, and ignition is withheld while the inducer works. Single-stage mainstream models (the R95P and (-)801S generation) carry one switch and report exactly these two codes. Communicating two-stage and modulating models carry two or three switches — low, high, and on modulating models a mid switch — and Rheem splits the reporting per switch, so the same two failures wear different numbers depending on which switch failed and at which firing rate. The proving logic, however, never changes: open when you must be open, closed when you must be closed, at the moment the control asks. A switch is therefore never condemned on one direction alone — a technician proves what the pressure at the switch actually was when it answered wrongly, because most wrong answers are the vent, the inducer or the hose telling the truth through a healthy switch.

Code 55: pressure switch stuck closed when it should be open

Rheem's chart definition preserves the timing that makes this code unique: the switch is still closed before the inducer runs — closed when it should be open, in the chart's own pairing of phrases. The furnace has not failed to make draft; it has refused to try, because its draft witness is already testifying before the event. Rheem's cause list for 55 is short and switch-side: a faulty (welded) switch, whose contacts have fused in the closed position; a switch physically bypassed in the field, whose 'closure' is a jumper wire rather than a diaphragm; loose or faulty wiring, which can present the control with a permanently made circuit; or abnormal negative pressure in the vent with the inducer off — wind or stack effect holding genuine suction on the switch at rest, the one cause on the list where the switch is honestly reporting a dishonest sky. The repair list follows the causes exactly: a technician replaces a welded switch, removes any bypass — and its reason, because a bypass was usually fitted to silence an earlier 57 and the vent story behind it is still unread — and checks venting and terminations for the at-rest pressure branch. Two owner disciplines matter here. First, recognition: a 55 arrives with silence — no inducer sound, no ignition sequence, just the code — and that silence is the sequence working, not a dead furnace. Second, honesty about history: if a 55 appears shortly after anyone 'fixed' a pressure-switch problem, the bypass branch moves to the top of the list and must be named when you call. On communicating multi-switch models, 55 names the HIGH switch specifically, and the display can pair it: both switches closed in the proving period displays 44 and 55 alternately — the low switch's stuck-closed code keeping time with the high switch's — which the family section below maps in full.

Code 57: pressure switch stuck open when it should be closed

Code 57 is 55's mirror: the switch will not close with the inducer running, so ignition cannot begin. Where 55 is met in silence, 57 is met in sound — the inducer runs, works, and keeps working, while the display holds its verdict that draft was never proved. Rheem's cause list for 57 is the vent-path census, and it should be read as a walk from the roofline back to the switch: a blocked or improperly terminated vent; vent length or elbows over the maximum the model allows; a faulty or disconnected inducer, the machine that makes the draft; damaged switch hoses, the tubes that carry the pressure signal the last few inches; wind gusts working against the termination; or, last, a faulty switch. The ordering is the economics of the code: terminations ice, nest and frost over; vent runs sag and pool condensate; hoses perish and split — while switches mostly report those failures faithfully. Owner work on a 57 is therefore real but outdoor: clear the vent and intake terminations of snow, ice, nests and debris, and confirm nothing has been built or planted over them. Everything past the termination — vent length accounting, hose replacement, inducer and switch proof with a draft measurement — is technician work, and Rheem's safety framing adds its absolute: never jumper the switch to prove the furnace 'really' works, because a jumpered switch converts the one witness to a blocked vent into a liar. On communicating models 57 names the HIGH switch with the inducer on high speed, and the furnace has a documented mercy for it: if the low switch stays closed, the furnace may drop to low fire — the display showing lower-case 'h' for low-stage heat, sometimes alternating with 57 — delivering reduced heat on the draft it could prove rather than none at all. A Rheem running weakly with a 57 in its history is usually telling exactly that story.

The wider family, mapped honestly: 44/45/46 low, 55/57 high, 65/67 name-only

On communicating Rheem models the pressure family widens, and this page maps it the way Rheem's manuals support — no further. The LOW pressure switch gets three codes of its own. Code 44: low switch closed with the inducer off — 55's failure at the low switch, with the same causes (faulty switch, field bypass, loose or faulty wiring, abnormal vent pressure at rest). Code 45: low switch open with the inducer on high speed — 57's failure at prepurge, where ignition cannot begin; its cause list is 57's census plus a faulty control (inducer relay), and its behaviour is patient: the inducer runs five minutes, then retries indefinitely. Code 46: low switch open with the inducer on low speed — seen only after heat is established and the furnace has switched to low fire, when the low switch drops out and the gas valve is de-energised; proving draft on high fire but losing it on low fire is its signature, and it needs a technician's draft measurement rather than a guessed switch. The HIGH switch keeps the mainstream numbers, 55 and 57, now read as high-switch reports. The displays pair the family truthfully: both switches closed in the proving period alternates 44 and 55; both switches open when both should be closed alternates 45 and 57. And then the honesty boundary: modulating (-)97V models carry a third, MID pressure switch, and Rheem's manual lists its codes as 65 (mid switch closed) and 67 (mid switch open). This site publishes those two numbers as names only. The two-stage (-)802V manual has no mid switch, and no second Rheem document checked in this pass details 65 or 67 — no definition beyond the listing, no causes, no fix — so none is invented here. If your modulating furnace shows 65 or 67, the mid pressure switch is the subject and the model's own manual is the authority; quote the code and model together when you call. A mapped gap is more useful than a fabricated meaning, because a fabricated meaning sends a technician — and a parts order — in a direction Rheem never pointed.

Low-fire fallback, paired displays, and the five-minute lockout

Three Rheem behaviours around these codes reward slow reading, because each converts a confusing display into a diagnosis. The first is the low-fire fallback already introduced under 57: on a communicating model whose high switch will not close but whose low switch holds, the furnace drops to low fire and heats at reduced capacity — lower-case 'h' on the display — rather than shutting down entirely. Owners experience it as a furnace that 'runs but cannot keep up' on the coldest nights, and without the code the complaint sounds like sizing or insulation; with the 57 in the history it is a high-draft proof failure, and the vent census (length, elbows, termination, hoses) is worked at high-fire severity, where the marginal restriction lives. The second behaviour is the paired display: Rheem alternates dual faults with the display's top segment lit briefly between them, so 44-with-55 and 45-with-57 are single events reported by two switches, not two independent failures arriving together — and a lone 57 with the low switch closed is a different (narrower) story from a 45-with-57, where no switch could prove anything and the vent or inducer itself is the first suspect. The third behaviour is timed: Rheem's lockout section documents a five-minute soft lockout when the low pressure switch fails to close after 60 seconds of continuous inducer operation at the start of a heat cycle — the display shows the appropriate fault, the lockout auto-resets after five minutes, and all lockouts can be cleared immediately, once the original fault is cleared, by cycling power, or (soft lockouts only) by cycling the heat call off for more than 2 seconds but less than 20 seconds. Code 45's own retry — inducer five minutes, retry indefinitely — is the same patience at the proving stage. Read together, these behaviours say the furnace distinguishes 'cannot prove at all' (escalating patience, then a pause) from 'can prove low but not high' (reduced fire) from 'proving device lying at rest' (no start at all) — three different faults that a one-line chart flattens into 'pressure switch problem', and three different first checks for the technician who reads them properly.

First checks in the field: stage, sound and terminations

A pressure code can be half-diagnosed before a single panel comes off, because the two codes announce themselves at different stages with different soundtracks, and the discipline is to catch the stage before resetting anything. Stage one is silence: kill the heat call, let everything stop, then call for heat and watch the first seconds. A furnace that shows its code without the inducer ever starting has failed the at-rest check — that is 55's territory (or 44's, for the low switch), and the suspects are the switch, its wiring and, on communicating models, the pairing on the display. Stage two is effort without ignition: the inducer starts and runs, the code holds, and no ignition sequence follows — that is 57's territory (or 45's at prepurge), and the suspects walk the vent census from the terminations inward. Stage three is partial success: the furnace lights and heats weakly, perhaps showing lower-case 'h', perhaps dropping out of high fire — the 57-with-low-switch-closed fallback or a 46 dropout story, where draft is marginal rather than absent and the restriction is partial. Between stages, the owner's one mechanical contribution is the terminations: go outside, find the vent and intake pipes, and clear snow, ice, frost, nests and debris, checking that nothing — a new fence, a stored bin, a drifted bank of snow — has been placed against them since the furnace last ran clean. Note the weather and the stage alongside the code (a 57 that appears only in high wind, a 55 that appears after a board change), because on this family the stage and the setting are worth more to the attending technician than any single part guess: they arrive knowing whether to bring a switch, a hose, a draft gauge or a ladder, and the furnace gets its first correct test instead of its third.

Shared causes: the vent-path census, worked in order

Codes 55 (at-rest branch aside) and 57 fail on the same physical chain, and Rheem's lists across the family name every link in the order a technician walks it. The terminations first: snow and ice after a storm, frost closing a screen over a cold snap, nests and debris in a mild season — the classic winter 57 and the one cause an owner can both find and fix. The vent run second: a blocked or improperly terminated run, and the arithmetic owners forget — vent length and elbows over the model's maximum, where every elbow spends draft like a length of pipe, so a vent extended during a renovation can cross the line from provable to unprovable without ever being 'blocked'. A sagging horizontal run belongs here too: pooled condensate narrows the pipe from inside and makes the fault weather-and-temperature dependent. The inducer third: faulty or disconnected, worn toward making almost-enough draft — the marginal inducer is why some 57s appear only on high fire or only on the longest runs. The signal path fourth: damaged, split, water-filled or disconnected switch hoses — cheap links that imitate every expensive one, inspected end to end before any switch is condemned. The weather fifth: wind gusts at the termination, which Rheem lists plainly; a 57 pattern that tracks windy nights is a termination-placement report. The switch last, on both Rheem lists — condemned on measured draft (pressure at the switch proven above its setpoint while its contacts sit in the wrong state), never on suspicion, and on 55's side the welded-contact and bypass branches are proved by the switch's state at rest with its wiring verified. Worked in that order, most pressure codes end at a termination, a hose or a vent correction — and the switch, when it truly is the switch, is replaced into a system already proven able to draft.

Never jumper the switch: the safety logic Rheem builds around these codes

Every code in this article exists because the furnace refuses to offer gas until its exhaust has somewhere to go — and the failure mode that proof guards against is the one that makes furnaces lethal: combustion gases, including carbon monoxide, spilling into the home from a blocked, disconnected or failed vent instead of travelling out of it. That single fact organises Rheem's whole design. It is why the control proves the switch in both directions — a switch stuck closed (55) blocks the cycle as firmly as one that will not close (57), because a device that cannot be caught lying in the safe direction cannot be trusted in the dangerous one. It is why a bypass fitted in the field is treated by Rheem's cause lists as a fault to be removed, never a repair to be kept. It is why a switch that drops out mid-run (46's story) de-energises the gas valve immediately rather than finishing the cycle on hope. And it is why 'adjusting' or jumpering a pressure switch to restore heat appears in no Rheem procedure: the adjustment deletes the evidence and the jumper deletes the witness, and a furnace that then fires into a restricted vent announces the danger in no way a household can sense — carbon monoxide is colourless and odourless, which is precisely why the proof is electrical and automatic rather than human. The correct responses are the ones in this article's order: clear the terminations, prove the vent and the inducer, renew the hoses, replace a switch that measurement condemns. If a Rheem furnace will not prove draft, it is not being stubborn; it is being the last working part of the safety system. Use backup heat, keep the household's CO detectors honest, and let a technician measure the draft. The switch is an inexpensive part. What it stands in for is not.

Key takeaways

• Listen before you look: 55 arrives in silence (the switch was already closed, so the inducer never starts), while 57 arrives with the inducer running and no ignition. The sound alone tells you which direction the switch failed.
• On communicating models the numbers name the switch: 44/45/46 are the low switch, 55/57 the high switch — and 65/67 on modulating models are listed by Rheem for the mid switch but not detailed in a second manual, so treat any fuller 'meaning' for them with suspicion.
• Never jumper, bypass or adjust a Rheem pressure switch to get heat — Rheem lists a field bypass as a cause of code 55, and the switch is the furnace's only proof that exhaust can leave the house.

Resetting the Rheem furnace

Rheem fault codes are shown as two-digit codes on the control's 7-segment display and clear once the fault condition is gone. Lockouts follow Rheem's documented rules: soft lockouts (the one-hour lockout, code 10, and the five-minute lockout when the low pressure switch fails to close after 60 seconds of inducer operation) auto-reset when their timer expires, and any lockout can be cleared immediately — once the original fault is cleared — by cycling power to the unit off and back on, or, for soft lockouts only, by cycling the heat call off for more than 2 seconds but less than 20 seconds. Hard lockouts are the exception: code 93 cannot be reset by cycling the heat call, and a dead-blower hard lockout (code 61, declared four times in one heat call) needs the repair plus a manual reset of power. Clearing a lockout without fixing its cause just restarts the count to the next one.

Frequently asked questions

Related guides

Facts on this page were checked against manufacturer documentation: pts.myrheem.com · www.manualslib.com · pts.myrheem.com · pts.myrheem.com · www.manualslib.com · my.ruud.com:443