Weigh-in method — the starting point for every new install
After evacuation to <500 µmHg and a passed decay test (rise <100 µmHg in 10 min), the nameplate charge is your baseline. On splits add a liquid-line correction: typically +20 g/m beyond the factory 5–7 m for R32/R410A, exactly per the unit's manual. Always meter liquid phase from an inverted cylinder, through a digital scale accurate to ±5 g, into the low side with the compressor running while throttling the valve — never charge R410A/R32 as vapour into the system, as that shifts the zeotropic blend ratio.
Weigh-in is the only reliable approach on systems with a TEV and a liquid receiver, and on R290 (R290 is dosed by weight only, with a charge limit set by standard — the charge is often just 150–500 g). The drawback: the scale cannot detect a leak or a wrong line length. It tells you how much you put in, not whether the system runs correctly.
Superheat — the check for capillary and fixed-orifice systems
Superheat is the difference between the actual suction-gas temperature and the saturation temperature read from suction pressure. On fixed-metering systems (capillary, orifice) it is the primary charge criterion: too little superheat = overcharge and flooding risk, too much = undercharge. Target is typically 5–8 K at the compressor suction, but read the target from the maker's method (charging chart by ambient temperature and evaporator entering-air conditions).
Measure with a digital manifold and a temperature clamp on the suction line right at the compressor, insulated. Account for the glide of zeotropic refrigerants (R407C, R448A, R454B) — use the dew-point saturation temperature for superheat. Add refrigerant in small increments and wait 10–15 min for stabilisation after each dose.
Subcooling — the check for TEV/EEV systems
On systems with an expansion valve (thermostatic or electronic) the valve controls superheat, so you verify the charge by subcooling on the liquid line. Subcooling = saturation temperature from discharge pressure minus the actual liquid temperature leaving the condenser. Target is typically 5–8 K (check the maker's charging chart); too little = undercharge with flash gas in the liquid line, too much = overcharge that floods the condenser and raises discharge pressure.
Conditions for a valid reading: stable steady state, clean condenser and evaporator, correct air/water flow. For zeotropes use the bubble-point saturation temperature for subcooling. As a cross-check, a correctly charged system shows a full sight glass with no bubbles — but the sight glass does not replace a subcooling measurement.
Combined procedure and common mistakes
In practice: weigh in to nameplate plus line correction, start the system, let it stabilise 15 min, then trim by superheat (capillary) or subcooling (TEV/EEV). Log pressures, temperatures and final charge in a record — this is also an F-gas requirement (EU Reg. 2024/573: logbook, leak checks, certified personnel). Never charge by feel or by the sight glass alone.
The most common errors: vapour-charging a zeotropic blend, reading superheat/subcooling without stabilisation or with a fouled coil, ignoring glide (using the wrong saturation temperature), and 'topping up' a leaking system instead of fixing the leak. If charge contamination is suspected — recover, evacuate and recharge with fresh refrigerant rather than correcting by addition.