Intermittent Mid-Cycle Lockout on a Condensing Furnace
A 90% AFUE two-stage condensing furnace with a hot surface igniter and a remote flame sensor locks out intermittently. Each call for heat begins normally: the inducer runs, the pressure switch closes, the igniter glows, the gas valve opens, and the burners light with a steady blue flame. The unit runs 6 to 9 minutes, then the gas valve drops out while the inducer keeps running; the board retries, and after three attempts it locks out. Flame current measured at the board is steady at 5.4 µA while the flame is stable, above the 1.0 µA minimum. The condensate trap is mounted on the left side. The drain line from the trap runs slightly uphill to a floor drain 8 ft away, and the trap contains visible rust-colored sediment. The vent is 2 in. PVC, 20 ft with two elbows, sloped back to the furnace. The house is tight and the indoor humidity is high. No fault code is stored beyond the generic lockout.
1. Explain how the condensate system could open the pressure switch mid-cycle, and why the failure appears after several minutes rather than at start-up.
Read the worked answer
In a condensing furnace the pressure switch proves that the inducer is moving flue gas, but it senses the pressure in the collector box/venting path. That path is also where condensate collects, and it is connected to the condensate trap.
With an uphill drain and a trap full of sediment, water cannot leave the trap as fast as it forms. During a call for heat, condensate accumulates in the collector box and trap until the water level rises enough to partially block the pressure switch port or the drain leg. The switch then sees reduced differential pressure and opens - even though the inducer is still running - so the board closes the gas valve. It does not happen at start-up because the trap starts relatively empty; it takes 6–9 minutes of run time for the water level to build. On the retry, the trap may still be partially flooded, which is why the retries fail and the board goes to lockout. The rust-colored sediment is consistent with a trap that has been holding water and corroding.
This also fits the generic lockout code: the board only records that the trial sequence failed, not which switch opened.
2. What tests and observations would confirm this diagnosis, and what would the readings look like?
Read the worked answer
Confirm by watching the sequence live rather than reading stored codes:
- Meter the pressure switch circuit while the unit runs - put leads across the switch (or from switch common to the board terminal) and watch for it to open at the moment the valve drops out. If 24 V appears across a closed switch, it has opened.
- Measure inducer pressure at the switch port with a digital manometer, before and at the moment of dropout. A reading that falls below the switch's setpoint (typically around −0.6 to −1.0 in. w.c. for a condensing unit, per the manufacturer) confirms a pressure problem rather than a board problem.
- Inspect the trap and drain - remove the trap, check for sediment and correct float/ball operation, confirm the drain line falls continuously to the termination, and confirm the termination is not submerged or in a frozen location.
- Check the vent for correct slope back to the furnace, sagging, or a low spot that collects water, and confirm the specified length and elbow count for 2 in. pipe.
- Verify the inducer wheel is clean and the collector box drains freely.
If the switch stays closed and pressure holds at dropout, the diagnosis shifts to flame sensing or the gas valve, not condensate.
3. What other faults produce this same mid-cycle dropout, and how do you rule each out?
Read the worked answer
Rule out the alternatives by watching which device changes state first at dropout:
- Flame sense dropout. 5.4 µA is well above the 1.0 µA minimum, so a steady reading argues against it - but soot or a micro-crack can make it intermittent as the burner heats. Rule out by watching microamp readings through the full run and cleaning or replacing the sensor and verifying the burner ground path.
- Gas valve or 24 V control dropout. Put a meter on the valve operator. If 24 V disappears at the same instant, the board or a safety opened the circuit; if 24 V stays and the valve closes, the valve is failing hot.
- High limit or rollout switch. A limit opening mid-cycle would also drop the valve. Check whether the limit circuit opens; if it does, suspect airflow (filter, coil, blower) rather than a limit defect.
- Vent pressure/recirculation. A tight house with high humidity and a 2 in. vent close to the maximum length can recirculate flue gas; check the termination clearances and combustion air provisions.
- Board or inducer thermal fault. Less common, but a failing inducer motor can slow after several minutes and drop pressure - verify with the manometer and by measuring inducer current.
Sequence of elimination: pressure switch first (since the trap is visibly suspect), then limit circuit, then flame sense, then the valve. Repair the trap slope and clean the trap, then run three complete cycles to confirm.