Complex fault diagnosis, repair and planned maintenance
A method beats a hunch
Use a consistent sequence: gather the complaint and history, verify the symptom yourself, compare against a known-good baseline or design data, form a hypothesis, test it with the least invasive measurement, then confirm the repair under load. Instrument error is a real cause - verify gauges, thermistors and clamp meters before condemning components.
Reading superheat and subcooling together
This pairing separates faults that look identical on suction pressure alone. Low superheat with high subcooling suggests overcharge or a liquid-line restriction. High superheat with low subcooling suggests undercharge or a starvation fault. High discharge temperature and pressure with normal charge suggests non-condensables or a dirty condenser. Low suction with normal superheat may simply be low load or low airflow, not a refrigerant fault.
Complex and intermittent faults
Consider control and sequencing faults, sensor drift, inverter and soft-start electronics, contactor and winding faults, phase imbalance, moisture/acid contamination, oil logging in traps and low legs, defrost termination faults, and BMS communication issues. For intermittent faults use data logging, trend analysis, thermal imaging and vibration checks rather than repeat visits.
Planned maintenance
Base PM on manufacturer intervals, duty, environment and F-Gas leak-check frequency. Record everything, close out actions, and feed repeat failures into a root-cause review.
Rule of thumb
Before replacing any component, confirm the measurement, confirm the load, and confirm the control is actually asking for the operation. 'Parts cannon' diagnosis is the classic exam trap - and the classic site failure.