Recurring Frozen Coil After Filter And Charge Corrected Decision Tree

Why this matters

A coil that freezes again after the obvious two causes (loaded filter, low charge) have already been corrected is a callback that costs you the relationship. The second visit means the first technician treated a symptom, not the root cause. Evaporator freezing reduces to one physics statement: the coil surface dropped below 32 F because heat absorption fell below the rate refrigerant could boil off the absorbed load. Filter and charge are two inputs to that balance, but a recurring freeze almost always means a third input was never measured. This tree walks the remaining airflow paths, the metering device, and the latent-load edge case that a one-and-done charge check will miss every time.

Symptom presentation

The customer reports ice on the suction line or the coil, water on the floor from a melted block, and weak or no cooling after a runtime of one to several hours. On arrival the coil may be fully iced or already thawed if the system was off. Key tell: the freeze recurs within hours to days of the prior repair, and the prior ticket shows a filter change, a charge adjustment, or both. The prior readings, if recorded, are your starting evidence. If subcooling and superheat were both brought into spec on the last visit and the coil still freezes, charge is not your problem.

Quick checks

Run these before tearing anything apart. Thaw the coil fully first by running the blower in FAN-only for two to four hours; pressures on an iced coil are invalid.

• Total external static pressure across the air handler with a manometer. Compare to nameplate rating (commonly 0.5 inWC).
• Blower amp draw against nameplate full-load amps.
• Return and supply temperature split at steady state, target 18 to 22 F at roughly 50 percent indoor RH.
• Filter again. A pleated high-MERV filter installed since the last visit can re-restrict in days.
• Evaporator coil air-side face for matted dust the prior tech did not pull the coil to inspect.
• Indoor wet bulb and outdoor dry bulb to confirm the unit is not running below its low-ambient cooling limit (most fixed-speed AC should not run cooling below 60 to 65 F outdoor).

Isolation tree

Branch A, static pressure high (above 0.7 inWC on 0.5-rated equipment): airflow is still restricted. Pull the filter and re-read. If static drops into range, the replacement filter is the culprit; spec a media cabinet or lower-MERV filter. If static stays high with filter out, the restriction is the coil air side, a collapsed flex return, an undersized return grille, or a closed zone damper. Trace each return path. A single closed-off return in a renovated room starves the whole system.

Branch B, static normal, blower amps low: the blower is moving less air than static implies because the wheel is loaded with dust or slipping on the shaft. Pull and inspect the blower wheel; caked vanes lose CFM with deceptively low static because the wheel never develops pressure. On an ECM, read commanded versus actual CFM at the controller; a failing ECM module derates without throwing static.

Branch C, static and amps normal, superheat low or zero at coil outlet: the metering device is overfeeding or the charge swung high after the prior adjustment. Re-verify subcooling on a TXV (target 8 to 15 F). Subcooling above 18 F with low superheat points to a TXV stuck open or an overcharge layered on top of the prior correction.

Branch D, static and amps normal, superheat high, subcooling low: the prior charge correction did not hold, meaning an active leak. A coil that was charged to spec and lost charge again in days has a leak rate that requires a search, not another top-off.

Branch E, all readings in spec but freeze recurs only on mild or humid nights: this is the low-load freeze. At low outdoor temperature or low sensible load with high latent load, suction pressure falls and coil temperature drops below 32 F even with correct airflow and charge. This is a control and application problem, not a fault.

Confirming diagnosis

Confirm airflow by CFM, not by feel. Target roughly 350 to 400 CFM per ton for cooling; below 300 CFM per ton the coil will frost under most conditions. Use the temperature-rise method on electric heat or the manufacturer's blower table cross-referenced to measured static to back into CFM. Confirm the metering device by watching superheat respond to a load change; a healthy TXV hunts then settles, a stuck valve does not respond. Confirm a leak with an electronic detector and bubbles at the suspected joint, then with a standing-pressure or nitrogen test if the leak is intermittent. Confirm the low-load case by logging suction saturation temperature against outdoor temperature over a full cycle; if SST tracks below 32 F only when outdoor drops below 65 F, you have a low-ambient freeze, not a fault.

Remediation

For airflow restrictions: clear the specific restriction found, then re-measure static to prove the fix. Right-size the filter and return grilles to the equipment's rated static. For a loaded blower wheel, clean or replace and verify amp draw returns to nameplate. For the metering device, replace a stuck TXV with the correct tonnage and bulb-mount it on the suction line at the 4 or 8 o'clock position with full insulation. For a recurring leak, perform a documented leak search and repair, evacuate to 500 microns, and weigh in the nameplate charge. For the low-load freeze, the fix is application-level: a low-ambient lockout, a fan-cycling control, or correcting an oversized system that short-cycles latent load. Do not keep adding refrigerant to chase a freeze that charge is not causing.

References

• AHRI Standard 210/240-2023, Performance Rating of Unitary Air-Conditioning and Air-Source Heat Pump Equipment
• ACCA Manual D, Residential Duct Systems (return-air sizing and static pressure)
• ASHRAE Handbook of Fundamentals, 2021, Chapter 1, Psychrometrics
• 40 CFR Part 82 Subpart F, EPA Section 608 (leak repair and venting prohibition)