Most off-grid failures can be diagnosed by tracing voltage from source to load with a digital multimeter. Energy stops moving somewhere between panels, controller, battery, and inverter—voltage readings at each node tell you which segment failed without replacing parts at random.
Use this as a practical field workflow, then compare findings with component manuals. For BMS threshold issues that mimic hardware faults, see Common BMS Configuration Mistakes.
Safety and scope before you probe
This guide assumes basic electrical safety awareness. It does not replace licensed electrical work where required.
Before testing:
- wear eye protection
- use insulated probes with correct category rating
- keep one hand away from live circuits where possible
- avoid probe-to-probe shorts on high-current battery terminals
- confirm the meter is on the correct DC V range before touching PV or battery
Fast diagnostic map: expected vs wrong readings
| Test point | Expected behavior | If reading is wrong | Likely fault zone |
|---|---|---|---|
| PV open-circuit at combiner | Near string Voc in sun | Very low or zero | Panel string, MC4, fuse, roof disconnect |
| Controller PV input | Similar trend to combiner | OK at combiner, absent at controller | Cable run, breaker, polarity |
| Controller battery output | Charging voltage when sun allows | No rise despite PV input | Controller config, MPPT fault, BMS block |
| Battery terminals under load | Stable within expected range | Deep sag on moderate load | Battery health, weak cell, cable resistance |
| Inverter DC input vs battery posts | Close match at same moment | Large delta under load | Undersized cable, bad crimp, fuse holder |
| Inverter AC output | Stable nominal AC | Dropouts with normal DC | Inverter overload, thermal, protection |
Step-by-step voltage tracing workflow
Step 1: Verify PV source first (open-circuit)
Measure string voltage before the controller with array disconnected from load (or at combiner). Compare to nameplate Voc × series count at current cell temperature—roughly within 10–15% in full sun. If wrong here, inspect panels, MC4 connectors, fuses, and roof disconnects before touching the battery.
Step 2: Confirm controller sees PV input
If PV is healthy at the combiner but absent at controller terminals, suspect the run between them—breakers off, reversed polarity protection tripped, or damaged cable.
Step 3: Validate charge path to battery
Measure at controller battery output and battery posts simultaneously during charge. A >0.3–0.5 V difference on a 12/24 V system under charge current suggests cable or terminal loss.
Step 4: Test battery behavior under real load
Static battery voltage can look fine while loaded voltage collapses. Test during appliance startup, not only at idle. Weak cells show disproportionate sag compared to bank history.
Step 5: Compare battery posts vs inverter DC input under surge
This is the highest-yield test for inverter shutdown mysteries. Measure both points during the same motor or coffee-maker start. Excessive drop points to resistance, not inverter "randomness."
What most people miss during meter testing
They test only at idle. Idle readings hide sag and intermittent failures that appear only during surge events.
They skip both sides of protective devices. Always measure before and after fuses and breakers. A fuse can look intact yet drop volts under amps.
They do not log measurements in sequence. Without timestamps and node labels, you misdiagnose where power disappeared. Write: time → location → V → load state.
They ignore BMS disconnect events. A BMS opening the contactor looks like a "dead inverter" until you measure battery side vs BMS output side.
They trust one sunny reading for a winter fault. Low sun masks PV problems that only appear at ** Voc threshold** on cloudy mornings.
Worked example: voltage drop exposes bad DC cabling
Scenario:
- 12 V system with 1,500 W inverter
- Coffee maker startup triggers inverter low-voltage alarm
Measurements during startup:
| Location | Voltage |
|---|---|
| Battery posts | 12.3 V |
| Inverter DC terminals | 10.6 V |
Voltage drop on path:
12.3 − 10.6 = 1.7 V
On a 12 V system under ~120 A surge (1,500 W ÷ 12 V), 1.7 V drop indicates excessive resistance—undersized cable, corroded lug, or heat-damaged fuse holder—not a "bad inverter."
Cable sanity (illustrative): AWG and length tables target under 3% drop at continuous load; surge events need heavier gauge or shorter runs than continuous math alone suggests.
Fix order: re-torque lugs → replace fuse holder → upgrade cable gauge → retest under same load.
Expected resting voltages for LiFePO4 (reference only)
| State (12 V nominal bank) | Approximate resting voltage |
|---|---|
| Nearly full | 13.2 – 13.6 V |
| Mid charge | 12.8 – 13.2 V |
| Moderate load territory | 12.4 – 12.8 V |
| Near empty (before cutoff) | 12.0 – 12.4 V |
Resting volts under load will read lower—always note load state in your log.
Practical field checklist
- Start at PV, then controller, then battery, then inverter—never skip upstream.
- Measure no-load and real-load conditions for each segment.
- Check both sides of every breaker and fuse under load.
- Compare battery-post voltage to inverter-input voltage during surge.
- Log readings with time, load, and sun conditions.
- Re-torque and re-test suspect connections before replacing hardware.
- If design—not fault—is the root cause, rerun loads in the WattSizing Calculator.
FAQs
Do I need an expensive meter for solar troubleshooting?
No. A reliable digital multimeter with correct DC voltage ranges and safe probe quality is enough for most workflows. A clamp meter helps for amp confirmation but is optional for voltage tracing.
Why does voltage look normal until I turn something on?
Load current exposes hidden resistance and weak components. Load testing beats idle readings.
Can a bad fuse holder cause intermittent inverter resets?
Yes. Corroded or heat-damaged holders create voltage drop that worsens under current spikes.
Should I test current (amps) too?
Current testing can help confirm charge rate or overload, but voltage-drop tracing usually finds faults faster with lower risk for many DIY users.
Is it safe to measure PV voltage in full sun?
Open-circuit DC voltage measurement is standard if you use proper DC-rated probes, correct meter range, and avoid touching other conductors. Do not short terminals with probes.
Why does my controller show PV watts but battery voltage never rises?
Suspect BMS charge block, cold-temperature charge inhibit, controller battery profile mismatch, or severe bank sulfation/degradation—measure controller output vs battery posts under charge.
Can a single weak cell mimic an inverter fault?
Yes. Under load, one weak cell pulls the whole bank voltage down while others look fine at rest—per-cell monitoring or BMS cell data helps confirm.
Next step: Once the fault path is proven, validate battery reserve, solar input, and inverter headroom in the WattSizing Calculator so the same failure does not return under the next heavy load.
