How to Charge a TXV System by Subcooling

Why this matters

A TXV system hides a wrong charge from anyone reading superheat. The valve's whole job is to hold evaporator superheat near its setting, so across a wide band of charge the suction line reads a steady, healthy number while the surplus or shortage shows up somewhere else. On a TXV system that somewhere else is the condenser: extra refrigerant backs up there as liquid, and a short charge leaves too little liquid to subcool. A tech who charges a TXV by superheat keeps adding until the valve finally loses control, which means a system running high head pressure, high amps and a hot compressor for the rest of its life.

Subcooling is the reading that moves with charge on a TXV or EEV system, which is why the maker publishes a target subcooling on the nameplate or in the installation instructions. The theory of both readings is in HVAC Superheat and Subcooling Method Reference. This card is the sequence at the condensing unit, in the order that keeps the number honest.

Safety before the hoses go on

R-410A runs a high side in the 300 to 400 psig range on a summer day, and liquid refrigerant at a hose fitting freezes skin on contact. Wear safety glasses and gloves rated for refrigerant work before you crack a port, use low-loss fittings, and keep your face out of line with a port while a hose is going on or coming off. Connecting gauges and adding or removing refrigerant on this equipment requires EPA Section 608 certification at the level the federal rule sets for high-pressure appliances; see related: EPA Section 608 Rules on a Residential Service Call.

The condensing unit carries 240 V at the contactor and a fan that starts without warning on a call. Read pressures and temperatures with every panel on and hands clear of the fan guard. If you have to open the electrical compartment, turn the unit off at the disconnect, lock it, and verify absence of voltage with a tester proved on a known live source before and after, per NFPA 70E Article 120 (120.5 in the 2021 edition, renumbered in 2024), with work practices under 29 CFR 1910.333(b)(2). The restore step is in step 8.

Where the readings come from

 CONDENSING UNIT, service valve side

 suction line, insulated ---- [SUCTION SERVICE VALVE]
                                 port -> low side hose
                                 read: dew point temp

 liquid line, bare copper --- [LIQUID SERVICE VALVE]
    |                            port -> high side hose
    |                            read: liquid pressure
    |                            -> bubble point temp
  clamp probe on the line set side, a few inches
  from the valve body, on clean copper, wrapped
  in insulation so sun and breeze do not reach it

 subcooling = bubble point temp - liquid line temp

The liquid pressure and the liquid line temperature have to be taken at the same place. A pressure at the condenser and a temperature at the indoor coil compare two different points with a line set of pressure drop between them, and the subcooling you calculate belongs to neither.

Steps

1. Confirm the metering device and the refrigerant. Pull the indoor coil's access panel or read its label and confirm it carries a TXV or an EEV, not a piston. A fixed orifice is charged by target superheat, and subcooling on a piston system does not tell you what this card assumes; see related: How to Charge a Fixed Orifice System by Target Superheat. Read the refrigerant off the outdoor nameplate and pick the matching pressure-temperature chart. A chart for the wrong refrigerant turns every reading after this into fiction.

2. Set airflow before you touch the charge. Clean filter, clean coil, the blower on the cooling speed the install calls for, and total external static within the rating plate. The reason is mechanical: low indoor airflow drops the load on the coil, the TXV throttles to hold its superheat, and the refrigerant it holds back piles up in the condenser. Subcooling rises. A tech who charges to target on a choked air handler removes refrigerant the system needs, and the day someone changes the filter the system is short. See related: Airflow Per Ton Targets and What Moves Them, and How to Estimate Blower Airflow From Static and the Fan Table.

3. Confirm you are inside the maker's charging window. The installation instructions state a minimum outdoor temperature for charging by subcooling, and often an indoor range. Below that outdoor floor the condenser runs so cold that head pressure collapses and the subcooling reading no longer tracks charge. If you are below it, stop here and weigh the charge in; see related: How to Charge a Winter Install When Superheat Cannot Be Trusted. Inverter and variable-speed systems usually have a charge or test mode that pins the compressor at a fixed speed; use it, because a target written for one speed is meaningless at another.

4. Run to steady state. Let the system run at least 15 minutes with all panels on. A common starting point for "steady": liquid pressure within 2 psig and liquid line temperature within 1 F across two readings five minutes apart. Reading sooner catches the condenser still filling, and the number drifts after you leave.

5. Read subcooling against the bubble point. Convert liquid pressure to saturation temperature using the bubble point column, then subtract the liquid line temperature. On R-410A the dew and bubble columns sit within a fraction of a degree, so either works. On a refrigerant with real glide (R-407C, R-454B and other blends) the dew column reads warmer than the bubble column at the same pressure, so using it overstates subcooling by roughly the glide, and a tech who believes it removes charge the system needs. See related: Refrigerant Glide, Dew Point and Bubble Point in the Field.

6. Adjust toward the nameplate target in weighed increments. Low subcooling means add; high means remove. Add blends as liquid from the cylinder, metered slowly into the suction port with the system running so it flashes before it reaches the compressor; dumping liquid in fast slugs the compressor. Remove charge into a recovery cylinder, never to atmosphere. Work in increments a few ounces at a time on a residential system, weigh every one, and wait at least 10 minutes after each before reading again. The waiting is the step people skip, and it is why charge "overshoots."

7. Confirm superheat shows a working valve. Subcooling on target proves the charge. It does not prove the TXV is feeding the coil properly. Read suction pressure, convert at the dew point, subtract from the suction line temperature near the service valve. A moderate superheat that held steady while you added charge says the valve is in control. Superheat near zero with subcooling on target says the valve is overfeeding (a loose bulb, a wrong valve). Superheat high with subcooling on target or high says a restriction or a starving valve, and adding charge will not fix it. Those branches belong to TXV Bulb Charge Loss Vs Superheat Adjustment Vs Replace Decision Matrix.

8. Restore and record. If the electrical compartment was opened, close it, remove your lock, restore the disconnect, and confirm the compressor and condenser fan start on a call. Remove hoses with low-loss fittings closed, cap the ports with their seals in place, and bubble-test both cores and caps. Write the ounces added or removed, the final pressures, subcooling, superheat, and both temperatures into the job record; see related: The Refrigerant Added and Leak Record SOP.

Worked example: a 3-ton TXV system that read fine on superheat

A 3-ton R-410A split, TXV on the indoor coil. The outdoor nameplate in this example lists target subcooling of 10 F, the manual gives a tolerance of plus or minus 2 F, and it sets 60 F outdoor as the floor for charging by subcooling. Conditions: 84 F outdoor, 76 F return air. The homeowner's complaint was a long run time on hot afternoons; the previous tech had "checked the charge" by superheat.

Airflow first. Total external static measured 0.48 in w.c. against a rating plate maximum of 0.50. The fan table at that static and the cooling tap gives about 1,150 CFM, which is 1,150 divided by 3, or about 383 CFM per ton. Airflow is not the problem.

First reading after 15 minutes and a stable second read:

• Liquid pressure 317 psig, bubble point 100 F on the R-410A chart
• Liquid line 96 F
• Subcooling: 100 minus 96 is 4 F, which is 6 F below target and outside the 2 F tolerance
• Suction 120 psig, dew point 41 F; suction line 60 F, so superheat 60 minus 41 is 19 F

Superheat of 19 F is high for a TXV, and low subcooling alongside it says the valve is wide open and still short of liquid. That is an undercharge, not a bad valve, and the subcooling is the reading that proves it.

First addition: 6 oz of liquid, metered over about three minutes, then a 10 minute wait. Liquid 330 psig, bubble point about 103 F; liquid line 96 F; subcooling 103 minus 96 is 7 F. Still 3 F low.

Second addition: another 6 oz, another 10 minute wait. Liquid 340 psig, bubble point 105 F; liquid line 95 F; subcooling 105 minus 95 is 10 F, on target. Suction 126 psig, dew point about 43 F; suction line 53 F; superheat 53 minus 43 is 10 F.

Total added: 6 plus 6, or 12 oz. Head pressure rose from 317 to 340 psig as the condenser filled, which is the direction it should move. Superheat fell from 19 F to 10 F as the valve regained control, which is step 7's proof that the TXV works. A TXV holds a near-normal superheat across a moderate shortage, which is how a superheat check can pass a system that is already short; by this visit the shortage had grown past the valve's range. A shortage in a sealed system also means a leak somewhere, so the ticket carried a leak search recommendation alongside the 12 oz.

Where the reading lies to you

A probe in the sun. An uninsulated clamp on a liquid line in afternoon sun reads warm. A warm liquid line reading makes subcooling look low, and the tech adds charge into a system that was right. The insulation wrap in the diagram is not optional.

A probe on painted or oxidized copper. Same direction as above, smaller error, and harder to spot. Clean the copper with an abrasive pad before clamping.

A wrong target. The nameplate target assumes the matched indoor coil the outdoor unit was rated with. If the coil was swapped for a different one, or the outdoor unit came off another install, the target on the label may not be the right one; check the installation instructions for the specific match before you charge to it.

A reading taken mid-cycle on a two-stage unit. First and second stage each have their own steady state. Charge in the stage the maker specifies, usually high stage, and confirm the call stays there for the whole sequence.

How to tell you got it right

Come back to the unit after the final reading, wait another five minutes, and read again. Subcooling within the maker's tolerance of target, superheat moderate and steady, liquid pressure stable within 2 psig, and the weights on the ticket adding up to the total change in the recovery or supply cylinder. If the second read wanders out of tolerance, the system was not at steady state when you stopped, and the last increment needs more time before anyone calls it done.

References

• Equipment maker's installation instructions and outdoor unit nameplate: target subcooling, tolerance, and charging temperature limits
• EPA Section 608 refrigerant management rules, 40 CFR Part 82 Subpart F
• NFPA 70E, Standard for Electrical Safety in the Workplace, current edition, Article 120; 29 CFR 1910.333(b)(2)
• See related: HVAC Superheat and Subcooling Method Reference
• See related: How to Charge a Winter Install When Superheat Cannot Be Trusted