What it is

Every inverter mini-split has self-diagnostics. When the board detects something outside its expected window — a sensor reading that makes no sense, a current spike, a lost conversation between indoor and outdoor — it stops and reports a code. Codes show up as a number/letter on the indoor display, a blink pattern on an LED, or on the handheld remote's screen.

The thing to understand: a fault code is a symptom the board noticed, not a diagnosis. It tells you which sensor or subsystem complained. Your job is to figure out why it complained. The specific code numbers vary by manufacturer (and I'm not reproducing anyone's table here), but the families of faults are universal, and once you know the families you can diagnose any brand.

How it works

The board constantly compares live readings against expected ranges. Thermistors should read plausible temperatures and track each other. Current should stay under a ceiling. The indoor and outdoor units should exchange data continuously over the comm wire. When any of those checks fails, the board latches a fault, often shuts down to protect itself, and flashes the code that maps to the failed check.

So the code points you at a category: a temperature sensor, the inverter/current path, the communication link, a fan, the EEV, or a pressure/protection limit. From there it's normal electrical and refrigeration troubleshooting.

In the field

Work a code in this order:

  1. Capture the exact code and where it displayed — indoor LED blink count, remote readout, or outdoor board LEDs. Photograph it.
  2. Look it up in THAT unit's service literature for what subsystem it maps to. Don't guess across brands — the same number means different things to different makers.
  3. Sort it into a family (below) so you know what you're hunting.
  4. Verify the obvious install causes first — especially on a recent install. Miswired comm, a thermistor knocked loose, a service valve left closed, low charge, or dirty/blocked airflow cause a huge share of codes.
  5. Measure, don't assume. If it's a thermistor family, ohm the sensor and compare to its temperature curve. If it's a current/inverter family, check charge, airflow, and the compressor windings before condemning the board.
  6. Clear and re-run after a repair to confirm the code stays gone under load, ideally in forced/test mode.

Normal values & targets

There's no single number table — codes are brand-specific — but these reference points help you work the families:

  • Thermistors are usually NTC type: resistance drops as temperature rises. Many indoor/coil sensors read in the roughly 5–50 kΩ at room temperature ballpark (varies by part). The test is comparison: a sensor reading wildly different from its sister sensor at the same temperature is the suspect.
  • Comm voltage: systems run a low-voltage DC or modulated signal on the comm leg; a dead or wrong reading there (vs the unit's spec) points to wiring, polarity, or a board.
  • Current/inverter faults trip near the unit's overcurrent ceiling — driven up by high head pressure (dirty condenser, overcharge, non-condensables) or a struggling compressor.
  • Steady operation itself isn't a fault — remember inverters run continuously at low speed by design.

Common faults & what they mean

Think in families:

  • Communication / serial error: indoor and outdoor stopped talking. Causes: comm wire reversed or wrong polarity, undersized or corroded wire on a long run, a loose terminal, water in a connection, or a failed board. Extremely common on fresh installs — almost always wiring.
  • Indoor/outdoor sensor (thermistor) fault: a temp sensor open, shorted, out of range, or popped off its clip so it reads air instead of coil. Ohm it and check placement.
  • High-current / IPM / inverter-drive fault: the power module saw too much current. Hunt high-head conditions first (dirty condenser coil, overcharge, restricted airflow, non-condensables), then compressor windings, then the board.
  • Fan motor fault: a DC fan motor not reaching commanded RPM — obstruction, failed motor, or board drive. The board watches feedback and faults if the fan won't spin to speed.
  • EEV / refrigerant-control fault: valve not stepping, coil unplugged, or a charge/restriction problem the board interprets as the valve failing.
  • Pressure / protection limit: high- or low-pressure or temperature protection tripped. Treat like any system trip — find the pressure problem (airflow, charge, condenser).

Tech tips & gotchas

  • A code is a clue, not a verdict. "Compressor fault" doesn't mean replace the compressor — it means the board saw a compressor-related limit. Low charge and dirty condensers throw "compressor" and "high pressure" codes on perfectly good compressors.
  • On a brand-new install, suspect the install first. Comm wiring, valve position, charge, and airflow cause most early codes. Don't condemn a board on day one.
  • Sister-sensor comparison beats memorizing curves. Two coil sensors at the same temperature should ohm close. The outlier is your bad one.
  • Loose or wet comm connections create intermittent faults that clear when you wiggle them. Re-terminate, don't just re-seat.
  • Megger and bench-check before you order a $700 board or a compressor. These big-ticket condemnations get made too fast off a single code.
  • Write the code and the fix in the customer file. Recurring codes across visits reveal patterns a single visit can't.

Safety / code notes

  • The inverter board's DC bus holds lethal voltage after disconnect — verify zero volts before probing the power section.
  • Restore disconnect and overcurrent protection to nameplate (NEC Article 440) after any repair.
  • On A2L equipment, certain protection codes tie into leak-detection/mitigation features — don't defeat them; follow the equipment's safety-system requirements.