Inverter and Variable-Speed Compressors
Modulating systems don't cycle on and off — they ramp compressor speed to match the load. That changes how they behave, how you read them, and why you charge them by weight instead of chasing a superheat number.
Chasing a weird inverter reading?
Walk the symptoms through Ventora's guided troubleshooting to rule out charge, airflow, and control faults in the right order.
In This Guide
What a Variable-Speed Compressor Actually Does
A conventional single-stage compressor has exactly two states: full output or off. The thermostat calls, the contactor pulls in, the compressor runs at a fixed speed until the space overshoots the setpoint, and then it shuts down. A two-stage compressor adds a second, lower capacity, but it's still a handful of discrete steps.
An inverter-driven compressor throws that model out. Instead of switching a fixed-speed motor on and off across the line, the outdoor unit contains a variable-frequency drive (the “inverter”) that takes incoming AC power, rectifies it to DC, and then synthesizes a new AC waveform at whatever frequency the control board commands. Change the frequency and you change the motor RPM; change the RPM and you change how much refrigerant the compressor pumps. Most residential inverter systems modulate roughly from 25–30% up to 100% of rated capacity, and premium units go lower.
The payoff is that the equipment can run at low, steady output for long periods and simply match the building's heat gain or loss. Long low-speed runtimes mean tighter temperature control, dramatically better dehumidification in cooling, and far fewer high-inrush starts — which is a big part of why these systems post the high SEER2 and HSPF2 numbers they do.
How Capacity Modulation Works
The compressor speed is only half the system. To keep the refrigerant circuit balanced across a wide range of speeds, nearly every inverter system pairs the variable-speed compressor with an electronic expansion valve (EEV) and a variable-speed (ECM) indoor blower. The control board reads suction and sometimes discharge pressure through transducers, reads coil and line temperatures through thermistors, and continuously trims the EEV to hold a target superheat while it ramps the compressor to hold the space temperature.
In practice, the sequence looks like this: a small load call brings the compressor up to a low frequency, the ECM blower delivers a matched CFM, and the EEV meters just enough refrigerant to keep the evaporator fed without flooding the compressor. As the load grows, the board raises the frequency, the blower speeds up, and the EEV opens further. Everything moves together.
Three parts move at once
- Compressor speed — sets total system capacity
- EEV position — meters refrigerant to hold superheat at any speed
- ECM blower CFM — keeps airflow matched to capacity (still targeting ~350–400 CFM/ton at full load)
Because those three are constantly self-adjusting, a snapshot of pressures and temperatures tells you far less than it would on a fixed-orifice, single-stage unit. The system you are metering may be at 40% one minute and 80% the next.
Why the Numbers Read Differently
On a standard system you lean hard on two measurements. For a TXV unit you set and verify charge by subcooling; for a fixed-orifice unit you use superheat. Those methods assume the compressor is running at one known speed and the metering device holds a fixed characteristic.
Neither assumption holds on an inverter system. If the compressor speed is floating and the EEV is actively driving superheat to a target, then reading superheat mostly tells you the EEV is doing its job — not whether the charge is correct. Subcooling shifts with speed too, because condenser mass flow and head pressure change as the compressor ramps. Take three readings ten minutes apart and you can get three different “answers.”
| Behavior | Single-Stage | Inverter / Variable-Speed |
|---|---|---|
| Capacity control | On / off | Continuous ~30–100% |
| Metering device | TXV or fixed orifice | Board-controlled EEV |
| Primary charge method | Subcooling or superheat | Weigh-in by data plate |
| Pressures at idle-to-mid load | Stable, predictable | Drift with commanded speed |
| Startup current | High LRA inrush | Soft-start ramp |
| Best verification | Gauges + delta-T | Forced-speed test mode + OEM chart |
Field caution
Do not condemn an inverter system's charge from a single set of gauge readings taken while it is free-running. Unless the board is locked into a fixed test speed, those numbers are a moving target. More than one good unit has been unnecessarily recovered and recharged because it was “read like a single-stage.”
Charging: Weigh In, Don't Guess
The manufacturer establishes the correct charge on the bench, then prints it on the data plate. Your job on a line-set install or a repair is to reproduce that charge by weight, then verify. The gauge readings become a confirmation step, not the setting method.
Follow the OEM procedure, but the shape is always:
Step 1 — Read the data plate. Record the factory charge, the length of line set it covers, and the adjustment per additional foot of liquid line. Note whether it uses an EEV or TXV.
Step 2 — Recover and evacuate. Recover per EPA 608, then pull below 500 microns and confirm the vacuum holds with the pump valved off. A wet or leaky system will never charge correctly no matter how precisely you weigh.
Step 3 — Calculate total charge. Factory charge plus line-set adjustment (see the worked example below).
Step 4 — Weigh in liquid. Use a calibrated scale. For blends like R-454B and R-410A, charge liquid from an inverted cylinder so the blend stays on-composition — pulling vapor fractionates the blend and shifts your charge.
Step 5 — Force full capacity. Put the board into its commissioning, charge, or test mode so the compressor holds a known fixed speed. This is what makes the readings meaningful.
Step 6 — Verify against the OEM chart. With the unit stabilized, compare subcooling, superheat, and pressures to the manufacturer's target for that speed and outdoor temperature.
Why weigh-in wins on these systems
A weighed charge is repeatable and speed-independent. Once the correct mass is in the system, the EEV and inverter manage the rest across the whole operating band. Verification then confirms you got it right instead of hunting a number that never sits still.
Ventora — Your AI HVAC Companion
Data-plate charge math, PT charts for R-454B and R-410A, and guided diagnostics — the AI HVAC assistant in your pocket. Free on iOS.
Worked Example: Total Charge Calculation
Problem
A 3-ton inverter heat pump lists a factory charge of 7 lb 8 oz (120 oz) good for the first 25 ft of line set. The data plate calls for 0.6 oz per foot of additional liquid line. Your actual line set is 55 ft. How much refrigerant do you weigh in?
Step 1: Find the extra line-set length
55 ft − 25 ft = 30 ft over the factory allowance
Step 2: Calculate the adjustment
30 ft × 0.6 oz/ft = 18 oz added
Step 3: Add to the factory charge
120 oz + 18 oz = 138 oz
Step 4: Convert to weigh-in target
138 oz ÷ 16 = 8 lb 10 oz total charge
Set your scale to zero with a full cylinder on it, then weigh out 8 lb 10 oz of liquid into the evacuated system. If the OEM specifies charge in grams, do the same math in grams — just keep the units consistent and never mix the factory value with a per-foot value in different units.
Don't forget the line set
Skipping the line-set adjustment is the most common charging mistake on these systems. On a long run, 18 oz of missing refrigerant on a 120 oz base is a 15% undercharge — enough to starve the evaporator, drive high superheat, and cripple low-ambient heating performance.
Field Diagnostics and Common Faults
Inverter systems are extremely reliable, but when they fault, the diagnosis lives as much in the control board as in the refrigerant circuit. Start by pulling the fault code — the board logs communication errors, drive faults, transducer faults, and EEV faults that a set of gauges will never show you.
- System runs only at low speed / never ramps up: often a communication fault between the thermostat, indoor board, and outdoor drive, or a demand signal the board isn't receiving. Verify the comm bus before touching refrigerant.
- Inverter / drive over-temperature trips: the drive's power module (IGBT) is heat-sinked and often refrigerant-cooled. A restricted condenser, low charge, or blocked cooling path can cook the drive and trip it offline.
- EEV hunting or stuck: erratic superheat and swinging suction pressure can point to an EEV losing steps or a failed coil, not a charge problem. Many boards let you command the valve open and closed to test it.
- Transducer drift: the board trims everything off pressure and temperature sensors. A drifting suction transducer makes a perfectly charged system behave badly. Compare the board's reported pressure to your manifold.
- Low-voltage or dirty power: VFDs are sensitive to supply voltage. Confirm voltage under load and check for loose lugs; brownouts can fault the drive.
Only after the board is happy and airflow is confirmed (clean filter and coil, correct ECM CFM) should you force full-speed operation and evaluate the charge against the OEM chart. Work the system in that order and you'll stop condemning good compressors.
A2L and Electrical Safety Notes
Most new variable-speed heat pumps sold since January 2025 use R-454B, an A2L (mildly flammable) refrigerant with a GWP of 466 that replaces R-410A. Its operating pressures sit within about 3% of R-410A, so the charging math and PT behavior feel familiar — but the handling rules are not. Respect the OEM charge limits, keep ignition sources away during service, use A2L-rated recovery equipment, and honor the integral leak-detection and mitigation the equipment relies on. You still need current EPA 608 certification to buy or handle the refrigerant.
The drive holds a charge — the electrical kind
The inverter's DC bus capacitors store a lethal voltage after the disconnect is pulled. Follow the manufacturer's discharge/wait procedure (commonly several minutes) and verify zero voltage before you probe the drive. Treat the outdoor board like an energized panel, not a contactor.
These units are the backbone of modern cold-climate heat pumps precisely because they can push capacity up at low outdoor temperatures. Charge them by the plate, verify at a forced speed, read the fault codes first, and respect both the A2L and the high-voltage drive — do that and variable-speed service becomes routine.
Charge and Diagnose With Confidence
Ventora is the AI HVAC assistant in your pocket — data-plate charge math, live PT charts, code lookups, and step-by-step troubleshooting for inverter and conventional systems alike. Start with a 7-day free trial.