Solar inverter showing an isolation / earth fault (often after rain)
The inverter refuses to start (or shuts down) on an isolation, ISO, ground-fault or Riso error — classically first thing on a dewy morning or after rain, sometimes clearing as the day dries out. Something on the DC side is leaking to earth.
Safety first
PV strings are live in daylight and this fault means DC is finding a path to earth somewhere. No live DC work: never open connectors or isolators under load, and treat frames, rails and roof metal as suspect until the leak is found. In Australia the PV DC side is accredited-solar territory — this guide is the diagnostic logic, not a licence.
Isolate, lock out / tag out, and prove dead before working unless a live test is specifically required, authorised, and carried out under proper supervision. Always follow local regulations, your site procedures, and the equipment manufacturer's documentation.
Apprentice mode
Full walk-through — the why, what readings mean, mistakes & safety.
Likely causes
Ranked from most to least likely.
- 1
Water in a rooftop DC isolator
Most likelyThe classic Australian failure — rooftop isolators fill with water, track to earth, and throw ISO faults that ease off as they dry. Many older models were the subject of recalls; treat any weathered one as suspect.
- 2
Damaged string cable insulation
#2Cable chafed on roof iron, tile edges or tray, cooked by heat, or chewed by vermin leaks to earth — worst when wet, which is why the fault loves mornings.
- 3
Water in PV plug connectors or a junction
#3Badly mated or non-matching plug pairs and unsealed junctions wick in moisture and track to the frame or roof.
- 4
A damaged panel leaking through its frame
#4Cracked backsheet or glass, or moisture inside a delaminating module, gives a leak path from the cells to the earthed frame.
- 5
Inverter measurement circuit at fault
Least likelyRare, but the Riso measurement itself can fail. Only conclude this after the array genuinely tests clean.
Reports are saved on this device to reflect what you actually find.
Testing sequence
Work through one test at a time. Expected reading and what each result means.
On the tools? Job Mode shows one test at a time with thumb-sized buttons — your place and readings carry straight over.
Enter Job ModeEstablish the pattern: does the fault appear on damp mornings and clear as the day warms up? Check the inverter log for times and (where shown) the measured isolation resistance.
Moisture-linked timing — faults at dawn/after rain that clear when dry — and a logged Riso value that dips below the inverter's threshold.
Moisture-dependent = almost certainly isolator, connector or damaged cable getting wet. Physical inspection will probably find it.
A hard fault in all weather is a solid leak — damaged cable or panel — and will show clearly on IR testing.
View all expected readings at once
Fault-finding flowchart
The same logic as a decision tree.
- 1start
Inverter ISO / earth-fault error
→ step 2 - 2decision
Does the fault come with damp (mornings/rain) and clear when dry?
Yes→ step 3No→ step 4 - 3decision
Visible water/tracking in a DC isolator, connector or junction?
Yes→ step 5No→ step 4 - 4decision
Does IR testing show one string low to earth?
Yes→ step 6No→ step 7 - 5result
Replace the wet component properly and seal the water entry — drying it isn't a repair.
- 6result
Halve-and-test along that string to corner the leak, then replace the section/component.
- 7result
Array tests clean in all conditions → suspect the inverter's measurement — manufacturer support.
Common mistakes apprentices make
- Restarting the inverter repeatedly until the fault 'clears' with the sun and calling it fixed — the leak is still there, waiting for rain.
- Opening DC connectors or isolators under load — DC arcs and doesn't forgive.
- Drying out a wet isolator and closing the lid instead of replacing it and fixing the water entry.
- Blaming the inverter's measurement before actually IR-testing the strings.
When to stop & escalate
All of the on-roof DC diagnosis and repair belongs with accredited solar people — bring one in rather than improvising. Panel-level leakage is a warranty conversation with the datasheet and your IR readings attached. If the array genuinely tests clean in all conditions and the unit still faults, it's a manufacturer job on the inverter.
If you're past your competence, authorisation, or the safe limits of the job — stop and hand it on. There's no fault worth getting hurt over.
Was this guide helpful?
Related faults
Solar inverter dead, blank, or showing a fault
The rooftop solar has stopped generating — the inverter display is blank, or it's alive but sitting on a fault code. The owner usually notices when the power bill jumps.
Solar inverter shutting down or derating on grid overvoltage
The inverter runs, then backs off or disconnects through the middle of the day with overvoltage events in its log. Generation (and the feed-in credit) suffers — worst on bright days when everyone's exporting.
RCD trips only in wet weather or after wash-down
The RCD holds fine when dry but trips after rain, washdown, or in damp conditions — pointing to moisture creating an earth-leakage path somewhere outdoors or in wet areas.
Intermittent fault that's hard to reproduce
Something fails occasionally — random trips, dropouts, or stoppages — but works fine when you go to look at it. The classic 'can't fault it on the bench' problem.