Furnace pressure switch stuck closed: what the control saw, and what it rules out
A closed pressure switch is reported before the inducer proves anything, so the flue advice written for an open switch does not apply. What Carrier, York, and Goodman each publish.
What this means
A stuck-closed report means the control found the switch contacts already made at a moment when combustion airflow could not have made them. Carrier calls it status code 23, pressure switch did not open, and defines it as the low or medium heat switch made when a call for heat is initiated. York flashes it as two red flashes, pressure switch closed with inducer pressure below the setpoint. The corrective actions all three manufacturers give point at the switch, its wiring, and the inducer, not at the flue.
Equipment and model context
Cross-brand symptom guidance
- Carrier 58MVC variable speed step-modulating condensing furnaces
- York Affinity YP9C modulating furnaces
- Goodman GMVC80 two-stage variable-speed furnaces
Each indication belongs to the furnace family it was published for. Status code 23 is a Carrier 58MVC entry, two red flashes is a York Affinity YP9C entry, and the wording used below for a low stage circuit closed at the start of a heating cycle is from the Goodman GMVC80 troubleshooting chart. The same flash count or the same two digits mean something else on another furnace, and a code read from a different manufacturer's table is not evidence about yours.
What you may notice
- The furnace will not light and the control shows a pressure switch code, while the inducer either never starts or runs without the sequence advancing.
- Heat works at low fire and the furnace refuses to step up, which Goodman lists as the symptom for the high stage version of the same entry.
- The code appears, clears itself, and returns later without anyone touching the equipment.
Likely causes
- Switch contacts that have welded, stuck, or drifted so they no longer open at rest, which is the condition each manufacturer names first.
- A short in the wiring between the switch and the control, which presents to the board as contacts that never open.
- Water or condensate held in the pressure tubing, which can hold a diaphragm displaced after the inducer has stopped.
- An inducer that is turning when the control expects it stopped, which puts real pressure on a switch that is working correctly.
Diagnostic paths
The inducer never starts and the code appears immediately
A call for heat produces a code and no motor noise at all, and the sequence stops before anything spins.
This is the case each manufacturer writes the entry for. Carrier defines status code 23 as the switch being made at the moment the call is initiated, and its own guide holds the code until the switch opens rather than locking the furnace out permanently. Record the exact indication and leave the electrical checks to a technician, because every published next step is inside an energized cabinet.
Low fire runs and the furnace will not go to high fire
The house heats slowly, the burner stays at minimum input, and a code is present or stored against the high stage.
Goodman lists exactly this symptom against its high stage entry: the furnace fails to operate on high stage while operating normally on low stage, with the inducer running. Note whether the shortfall appears only in cold weather, since a furnace that never needs high fire in mild weather can carry the condition unnoticed for weeks.
The code appears and clears without intervention
Heating works for days, then a cycle fails and the next one succeeds, with no pattern anyone has found.
Intermittent behaviour fits a switch drifting near its setpoint, a wiring short that moves, or moisture in the tubing, and Carrier's design allows it: the control flashes the code until the switch opens and then begins the cycle. Log the date, the time, and the outdoor conditions for each occurrence, because a pattern tied to weather points somewhere different from one tied to time since the last run.
| Reported condition | What the control observed | Where the manufacturer sends the technician |
|---|---|---|
| Switch did not close | Contacts still open once the inducer should have proved flow | Flue and inlet piping, the drain system, the tubing, and inducer performance |
| Switch did not open | Contacts already made before the inducer could have made them | The switch itself, its wiring for a short, and whether the inducer is running |
| Furnace family | Indication | Published wording |
|---|---|---|
| Carrier 58MVC | Status code 23 | Pressure switch did not open, defined as LPS or MPS made when a call for heat is initiated |
| York Affinity YP9C | Two red flashes | Pressure switch closed with inducer pressure below setpoint, so closed when it should be open |
| Goodman GMVC80 | Troubleshooting chart entry | Low stage pressure switch circuit closed at start of heating cycle, with contacts sticking listed beside it |
What the control is testing, and when
A combustion pressure switch is a normally-open device that closes once the inducer pulls enough negative pressure to prove flow. The control reads it twice for different reasons. Before the inducer starts it expects the contacts open, because nothing has yet proved anything. After the inducer has run it expects them closed, because flow should now exist.
The stuck-closed report comes from the first of those reads. Carrier states the definition without ambiguity for the 58MVC: status code 23 indicates the low or medium heat pressure switch is made when a call for heat is initiated. York describes the same thing from the pressure side, calling it a switch closed with inducer pressure below the switch setpoint, and adds the plain-language gloss that the switch is closed when it should be open.
That timing is the whole distinction. A blocked flue cannot make a switch close, because a blockage removes flow rather than creating it. Whatever closed those contacts did so without the airflow the switch exists to sense, which leaves the switch, the circuit around it, or an inducer that is running when it should not be.
What the manufacturers actually send a technician to check
Goodman's troubleshooting chart is the clearest of the three, because it prints both conditions side by side with different remedies. Against the low stage circuit closed at the start of a heating cycle it lists sticking contacts and a short in the pressure switch circuit wiring. Its corrective actions there are to verify induced draft blower performance, replace the switch, repair the short, and inspect the hose. Against the low stage circuit that is not closed it lists blocked flue or inlet air pipe, a blocked drain system, or a weak induced draft blower.
Carrier's sequence for status code 23 follows the same logic in a different form. It asks whether the inducer is running, and if it is, sends the technician to the inducer drive signal rather than to the switch. If it is not, the checks run along the switch wiring and the control connector, ending at rewiring, replacing the pressure switch assembly, or replacing the control.
York gives the shortest instruction of the three against two red flashes: check pressure switch. Read together, the three documents agree on where the answer is not. None of them opens the closed condition with a vent inspection.
Why the common advice is written for the other fault
The advice a search returns for a furnace pressure switch is written for the open condition, and for that condition it is sound. A vent terminal packed with snow, a nest at the intake, or a condensate trap holding water all reduce the draft the switch needs, so the switch stays open and the furnace refuses to light. That is a real sequence with a real remedy.
Applied to a closed report the same advice sends a homeowner outdoors to clear a termination that was never the problem, and sends a technician to a vent system the control has said nothing about. Worse, it invites the conclusion that the furnace is unsafe when the report is about a switch that has stopped opening.
The safety position does not change either way. A pressure switch is a combustion safety device, and a furnace that is reporting one has withheld gas heat on purpose. Nothing on this page argues for running it, and nothing here is an instruction to bridge, tap, or bypass a switch to get through a cold night.
What an owner can establish without opening anything
Three observations are worth more to the visit than any guess about the component. The first is whether the inducer runs at all during a failed call, because that single fact splits Carrier's published sequence in two. The second is whether heat is refused entirely or only at the higher input, since Goodman ties the high stage version to a furnace that runs normally on low stage. The third is whether the fault repeats every cycle or comes and goes.
The exact indication matters as much as the symptom. Two red flashes on a York and status code 23 on a Carrier describe the same underlying condition, but neither number transfers to the other cabinet. A flash count copied from a general article is not evidence about the furnace in the basement. Photograph the label inside the door alongside the flashing indicator.
What to give the technician
Hand over the full furnace model number, the indication exactly as it appeared and where it was displayed, whether the inducer ran during the failed call, whether low fire still works, and the log of when the fault appeared. That set tells the technician which branch of the manufacturer's own sequence to start on before the panel comes off.
Safe homeowner checks
- Read the model number from the rating plate and photograph the diagnostic label printed inside the furnace door, which is where the flash or code key for that furnace is published.
- Watch a failed call from outside the cabinet and note whether the inducer motor runs, does not run, or runs and keeps running after the call ends.
- Record whether heat is refused completely or only at higher input, and whether the failure repeats on every cycle.
- Log the date, time, and outdoor weather for each occurrence, since an intermittent fault is diagnosed from its pattern rather than from a single event.
Is a reset appropriate?
Cutting power clears the display and any stored history, which on an intermittent fault removes the only record of how many times it happened. Carrier's control holds this code until the switch opens rather than latching it out, so a furnace that starts after a power cycle has not been repaired by one.
When to call a technician
- Every corrective action these manufacturers publish for the closed condition sits inside an energized cabinet, including measuring at the switch terminals and at the control connector.
- Removing a wire from a pressure switch, tapping the switch body, or jumpering a switch to force a start defeats a combustion safety device and is not owner work.
- Gas heating equipment that is withholding heat on a safety input needs a qualified technician, and in the United Kingdom gas work is restricted to a Gas Safe registered engineer.
Questions people ask about this
Does a stuck closed switch mean my flue is blocked?
No. A blockage removes airflow, and this report says the contacts were already made before airflow could have made them. Goodman lists blocked flue and inlet piping against the opposite entry, the circuit that is not closed, and gives the closed entry a separate set of corrective actions.
Why does my furnace run on low fire but not high?
Goodman lists that as the symptom for the high stage version of the same condition: the furnace fails to operate on high stage while operating normally on low stage, with the inducer running. Its corrective actions are to diagnose and replace the high stage switch, or to repair a short in the wiring.
Can I tap the switch to get heat back tonight?
No. A pressure switch is a combustion safety device, and forcing it to change state removes the proof of draft the furnace lights on. The manufacturers that describe a switch as sticking treat that as grounds for replacement, not as a temporary fix.
Is this always a failed switch?
The published entries name more than one cause. Goodman lists sticking contacts and a short in the pressure switch circuit wiring, and Carrier's sequence branches to the inducer drive signal when the inducer is found running. A stuck-closed report identifies the circuit, not the part.
Why does the code sometimes clear on its own?
Carrier's control flashes status code 23 until the switch opens and then begins the cycle, so an intermittent condition can produce a failed call followed by a normal one. That behaviour is documented rather than a sign the fault has resolved.
Evidence recordHow this page was checked
service manual · checked 2026-09-07
Every technical claim above was written from primary documentation held in the HVAC Bench evidence record: Carrier, York and Goodman technical literature. Where a source limits a definition to certain models, test conditions, or product classes, that limit is repeated here rather than generalised.
- Documentation class
- service manual
- Scope of the definition
- Confirm against the exact model manual
- Last checked
- 2026-09-07