Temperature split is often the first number a tech takes on a cooling call, because it only needs a thermometer and two holes in the ductwork. It won't tell you what's wrong, but it quickly tells you whether the system is moving heat the way it should and whether to look at airflow or the refrigerant side next.
Temperature split in plain English
Air goes into the indoor coil warm and comes out cold. The difference is the split. DOE's Building America measure guideline defines it as the difference between return air and supply air dry-bulb temperatures, measured upstream of the blower and downstream of the cooling coil1.
Formula
A cooling system does two jobs: lowering the air temperature (sensible cooling) and pulling moisture out (latent cooling). The split only shows the first job. On a humid day more of the capacity goes to moisture, so the split gets smaller even when the system is working perfectly. That's why the target moves.
How do you measure temperature split?
- 1
Check airflow first
DOE's guide says the split only indicates a charge problem once airflow is verified above 300 CFM per ton. Check the filter, blower and static pressure.
- 2
Run the system 15 minutes
Let temperatures settle before reading, the same rule DOE's guide uses for charge tests.
- 3
Read return air
Dry bulb and wet bulb in the return, upstream of the blower (or at the return grille for wet bulb), probe centered in the airstream.
- 4
Read supply air
Dry bulb in the supply, downstream of the cooling coil, probe centered in the airstream.
- 5
Subtract and compare
Return minus supply is the split. Look up the target for the measured return dry bulb and wet bulb.
Worked example (example numbers)
| Reading | Value |
|---|---|
| Return air dry bulb | 78°F |
| Return air wet bulb | 63°F |
| Supply air dry bulb | 58°F |
| Measured split | 78 − 58 = 20°F |
| Target from the split table for 78°F db / 63°F wb (example) | 19°F |
| Difference | +1°F: within ±3°F |
If supply had read 63°F, the split would be 15°F, 4°F under the example target, and DOE's guide would call that a capacity problem worth diagnosing1.
Common mistakes
- Reading the supply before the air has mixed, for example right at one corner of the coil.
- Ignoring wet bulb. A split without the return wet bulb can't be judged.
- Return leaks. DOE's guide notes duct leakage can lead to incorrect split readings, especially when the return pulls attic air1.
- Skipping the airflow check and adding refrigerant to fix what is really a dirty filter.
What is the target temperature split?
The target comes from a table indexed by return air dry bulb and wet bulb. DOE's guide uses one developed by Carrier and modified by the California Energy Commission for Title 24, and says the measured split should be within ±3°F of the table value1.
For design, DOE's guide summarizes ACCA Manual S guidance: at a sensible heat ratio below 0.80, airflow is chosen for a supply temperature about 21°F below return; above 0.85, the split is reduced to about 17°F by increasing airflow1. ACCA's blog shows the same trade-off from the other side: the same 2.5-ton unit delivered more sensible cooling at 450 CFM per ton than at 3502.
There isn't one universal "good" split, so treat any single number quoted without conditions with suspicion.
Why does humidity change the target?
A cooling coil's total capacity is shared between sensible cooling, which lowers the air temperature, and latent cooling, which condenses moisture. The split only measures the sensible part. ACCA's blog shows how much that share can move on one unit: a 2.5-ton system at 63°F entering wet bulb put 52% of its capacity into sensible cooling at 350 CFM per ton and 61% at 450 CFM per ton, while total capacity barely changed2.
So two healthy systems, or the same system on two different days, can show different splits. A humid house pushes more capacity into moisture removal and the split shrinks. A dry house does the opposite. Reading the split against return wet bulb is what turns it from a guess into a test.
What causes a high or low split?
| Split vs target | Likely causes | Next check |
|---|---|---|
| More than 3°F high | Low airflow: dirty filter, slow blower, dirty coil, restrictive return | Static pressure and airflow |
| More than 3°F low | Lost capacity: undercharge, liquid line restriction, compressor or metering problem | Superheat and subcooling |
| On target | Probably fine, but not proof | DOE says confirm with charge tests: overcharge and non-condensables can still show a normal split1 |
Temperature split vs superheat vs temperature rise
| What it measures | Where it's used | |
|---|---|---|
| Temperature split (cooling) | Air temperature drop across the indoor coil | Quick check of cooling performance and airflow |
| Superheat and subcooling | Refrigerant temperatures vs saturation | Setting and verifying refrigerant charge |
| Temperature rise (heating) | Air temperature increase across a furnace or heat strips | Checking furnace airflow against the data plate range; estimating CFM |
In heating, the same subtraction runs the other way: supply minus return. With a known heat input, the rise also gives you airflow through the sensible heat formula, CFM = BTU/h ÷ (1.08 × rise)3. See CFM.
Free calculatorSuperheat, subcooling & target superheat calculatorEnter pressures and line temperatures; get superheat, subcooling and a target superheat check.How temperature split shows up in TradesWork
The TradesWork technician app includes temperature split and temperature rise calculators that work offline, and split readings can be saved to the equipment record alongside superheat and subcooling. A maintenance form can also require the readings, with units and sanity bounds, before a tune-up closes.
Footnotes
-
D. Springer and B. Dakin, Measure Guideline: Air Conditioner Diagnostics, Maintenance, and Replacement, U.S. DOE Building America, DOE/GO-102013-3766, March 2013, pp. 8, 14–15, 25–30. https://www1.eere.energy.gov/buildings/publications/pdfs/building_america/measure_guide_air_cond_diagnostics.pdf ↩ ↩2 ↩3 ↩4 ↩5 ↩6
-
M. Akins, "400 CFM Per Ton, Or Is It?" ACCA HVAC Blog, Feb. 4, 2021. https://hvac-blog.acca.org/400-cfm-per-ton-or-is-it/ ↩ ↩2
-
B. Johnson, "Btu Buddy 159: Heat Pump Air Distribution Problem," ACHR News, June 20, 2016 (trade press). https://achrnews.com/articles/132765-btu-buddy-159-heat-pump-air-distribution-problem ↩
