What subcooling actually measures
Refrigerant leaves the compressor as hot vapour and enters the condenser, where it gives up heat to outdoor air. Partway through the coil it reaches its saturation temperature and starts to condense; from there on, vapour and liquid coexist at the same temperature while the phase change completes. Once the last of the vapour has condensed, any further heat rejection has nothing left to condense, so it lowers the temperature of the liquid instead. Subcooling is that final temperature drop — how many degrees below the boiling point the liquid has been taken.
You cannot read it off a thermometer alone, because "the boiling point" depends on pressure. So you measure both: pressure tells you what temperature the refrigerant would boil at, and a clamp-on probe tells you what temperature it is actually at. The difference is subcooling.
The reason it matters for charging is geometric. Liquid refrigerant occupies the bottom of the condenser. Add refrigerant to a system and it backs up into more of the coil, giving the liquid more coil surface to sit against and more time to cool — subcooling rises. Remove refrigerant and the liquid line retreats, subcooling falls. On a system with a thermostatic or electronic expansion valve, which meters flow to hold evaporator superheat constant regardless of charge, subcooling is the only clean indicator of how much refrigerant is in the system.
Reading the pressure-temperature relationship
The arithmetic is one subtraction. The work is in the first term.
Subcooling = T_sat(P_liquid) − T_liquid line
Every refrigerant has a fixed relationship between saturation pressure and saturation temperature — that is what a P/T chart is, and what the scales on a manifold gauge print. R-410A at 317 psig is saturated at 100 °F. R-22 at 196 psig is saturated at 100 °F. R-134a reaches 100 °F at only 124 psig. The temperature is the same; the pressures are entirely different, which is why the refrigerant selection above changes everything.
Two details separate a correct reading from a plausible one.
Use the bubble point for blends. Zeotropic blends do not condense at a single temperature — they have a glide, with the dew point at the vapour end and the bubble point at the liquid end. Subcooling is measured where the refrigerant is fully liquid, so the bubble point is the correct column. R-410A is a near-azeotrope with a glide well under a degree, so the distinction barely shows; on a blend such as R-407C, with several degrees of glide, using the wrong column puts your answer out by that much.
Measure pressure and temperature at the same place. Liquid line pressure at the outdoor unit and liquid line temperature at the indoor coil are not a matched pair: the line and the filter drier between them lose pressure, and a long line set in a hot attic gains heat. Put both readings within a few inches of each other at the condensing unit service port unless the manufacturer says otherwise.
Worked example: an R-410A system on an 85 °F day
You are checking a TXV-equipped R-410A split system. The charging label calls for 10 °F of subcooling. Outdoor air is 85 °F. After 15 minutes of run time you read 317 psig at the liquid port and 90.0 °F on a clamp probe six inches away.
- Convert pressure to saturation temperature. On the R-410A table, 317 psig corresponds to 100 °F.
- Subtract. 100.0 − 90.0 = 10.0 °F of subcooling.
- Compare to target. 10.0 − 10.0 = 0. The charge is correct.
- Check the condenser split. 100.0 − 85.0 = 15.0 °F, which is a normal split for modern high-efficiency equipment.
Now change one thing. Suppose the coil is packed with cottonwood seed and the condensing temperature climbs to 120 °F while the liquid line reads 106 °F. Subcooling is 14 °F — four degrees over target, which looks like an overcharge. But the condenser split is 120 − 85 = 35 °F, far wider than it should be. The refrigerant is not the problem; the coil is. Recovering refrigerant here would leave you undercharged the moment somebody cleans the condenser.
That pairing is the whole reason the split appears on this page beside the subcooling figure.
How to read the number
The target comes from the equipment, not from a rule of thumb. Most residential split systems land somewhere between 8 and 14 °F, and the charging label on the condensing unit gives the exact figure — often as a table that varies with outdoor temperature or indoor wet bulb. Use that. Where no label survives, the installation instructions for the model are the next authority, and the distributor after that.
Below target on a TXV or EEV system means there is not enough refrigerant to keep the bottom of the condenser flooded. Before you add any, find out where the missing refrigerant went: a system that was correct last season and is low now has a leak, and topping it up without repair is both bad practice and, for most refrigerants, a regulatory problem.
Above target has four common causes and only one of them is overcharge. Restricted condenser airflow, a fouled coil, a condenser fan running slow or backwards, and non-condensable gas trapped in the system all raise condensing temperature, and all of them raise subcooling with it. The condenser split separates them: overcharge raises subcooling with a normal split, while every airflow fault raises subcooling and widens the split together.
Near zero or negative means there is no liquid seal at all. Either the system is severely short of refrigerant, or something upstream of the measurement is restricting flow and flashing the liquid early. A negative reading is not a small error; it is a system that cannot feed its metering device.
Why a wider split means a heat-rejection problem is straightforward. The condenser rejects heat at a rate set by Q = U × A × ΔT. For a given amount of heat, anything that reduces the effective coil area or the air-side heat transfer coefficient must be paid for with a larger temperature difference. That is also why a physically larger, higher-efficiency condenser runs a narrower split than an old small-coil unit rejecting the same heat.
Pressure-temperature reference for common refrigerants
| Saturation temp (°F) | R-22 (psig) | R-410A (psig) | R-134a (psig) |
|---|---|---|---|
| 40 | 68.5 | 118.2 | 35.0 |
| 50 | 84.0 | 143.1 | 45.4 |
| 60 | 101.6 | 169.9 | 57.5 |
| 70 | 121.4 | 201.1 | 71.1 |
| 80 | 143.6 | 235.1 | 86.8 |
| 90 | 168.4 | 273.9 | 104.3 |
| 100 | 195.9 | 317.4 | 124.3 |
| 110 | 226.4 | 364.9 | 146.4 |
| 120 | 259.9 | 418.3 | 171.2 |
| 130 | 296.8 | 478.0 | 198.7 |
Interpolate between rows on a logarithmic pressure scale, which is what the calculator above does. A digital manifold with the refrigerant profile loaded does this for you and is worth the money for A2L work.
Mistakes that produce a wrong subcooling figure
- Charging by subcooling on a fixed-orifice system. A piston or capillary system has no valve modulating flow, so its subcooling wanders with indoor and outdoor conditions. Those systems are charged by target superheat instead.
- Reading the dew-point column on a blend. Subcooling is a liquid-side measurement and needs the bubble point. On R-410A the difference is fractions of a degree; on a high-glide blend it is several degrees of pure error.
- A probe that is reading air. Clamp on clean, bare copper and insulate the probe from the breeze. On a hot day an uninsulated clamp on a 90 °F line in 100 °F air reads several degrees high, which understates subcooling.
- Charging before the system is stable. Give the system 15 minutes of continuous run time. Readings taken in the first few minutes reflect refrigerant migration, not the operating charge.
- Charging in mild weather. Below roughly 60 °F outdoors, head pressure falls, the condenser floods on its own, and both subcooling and superheat lose meaning. Weigh the charge in instead, using the nameplate charge plus the line length adjustment.
- Ignoring the filter drier. A partially plugged drier drops pressure between the condenser and your measurement point. If subcooling looks fine at the condenser but the system behaves as though it is starved, check the temperature difference across the drier.
Manufacturer data always wins
Every number on this page is a general engineering relationship. The target subcooling for your specific condensing unit, at the specific outdoor temperature you are working in, is on the equipment charging label and in the installation instructions, and it is the only figure with authority. Some manufacturers publish a single value, some a table indexed by outdoor temperature, and some require charging by weight with subcooling used only as a verification check. Read the label before you open a cylinder.
Subcooling, superheat, and the newer refrigerants
Subcooling and superheat are the two halves of the same diagnostic. Superheat is measured on the low side and tells you how well the evaporator is being fed; subcooling is measured on the high side and tells you how much refrigerant the system holds. Reading them together separates faults that either one alone leaves ambiguous — high superheat with low subcooling is the classic undercharge signature, while high superheat with high subcooling points at a restriction between the two.
Which of them you charge by depends on the metering device. A TXV or EEV holds superheat roughly constant by design, so superheat carries no information about charge and subcooling does all the work. A fixed orifice has no such feedback, so superheat moves with charge and becomes the charging variable, compared against the target from indoor wet bulb and outdoor dry bulb.
The refrigerant transition adds a practical wrinkle. R-454B and R-32 have replaced R-410A in new residential equipment under the AIM Act phasedown, and their saturation pressures are close to R-410A's but not identical — close enough that using an R-410A scale would produce a small, consistent error rather than an obvious one. Use a digital manifold with the correct refrigerant profile loaded, or the manufacturer's own P/T chart, and enter the saturation temperature directly in the option above. The arithmetic on this page does not change; only the first term does.
Once you have both readings, the compression ratio calculator turns the same two pressures into a check on how hard the compressor is working, and the delta-T calculator confirms whether the air side agrees with what the refrigerant side is telling you.