Low Solar Output: Troubleshooting Checklist (Common Causes)

Quick answer

Low output almost always comes down to one of five causes, in this order of likelihood:

1. Soiling — dirt, pollen, bird droppings, or snow on the glass.
2. Shading or season — new shade from tree growth, or you’re comparing December to July.
3. Wiring and connections — a loose MC4, corroded lug, or tripped breaker acting like a resistor.
4. Charge controller or inverter settings — wrong battery profile, a derating error code, or float mode that just looks like a fault.
5. Real degradation — panels lose capacity slowly; sudden drops are almost never this.

The fix sequence is the same order: check what’s cheap and visible first, do the voltage/current math second, and only then start disconnecting things. Most “my system got weak” reports end at steps 1 or 2 — and a surprising number aren’t a fault at all, just winter.

Key takeaways

• Establish a baseline before diagnosing. Nameplate watts are lab numbers; a healthy array delivers roughly 70–85% of nameplate in good conditions after real-world losses.
• Compare like with like: this month vs the same month last year, not vs your summer peak.
• Shading is disproportionate. A small shadow on a series string can cut array output 20–40%, far more than the shaded area suggests.
• A multimeter plus the spec sheet answers most questions in 20 minutes — panel voltage and current measurements separate “input problem” from “equipment problem.”
• Sudden, large drops point to wiring, shading, or an error code. Slow, gradual decline over years points to soiling habits or normal degradation.

Establish your baseline first

You can’t diagnose “low” without defining “normal.” Start with nameplate and derate it honestly.

A panel’s rated watts (STC) are measured at 25°C cell temperature under standard lab light. Real installations run hotter and lose a few percent in wiring, connections, and the inverter. A conservative working derate for a well-installed grid-tied array is 0.75–0.85; dirty or hot rooftop arrays sit at the low end.

Expected daily Wh ≈ Array watts × Peak sun hours × System derate

Worked example — 6 × 400W panels (2,400W array), 4.5 peak sun hours, 0.80 derate:

2,400 × 4.5 × 0.80 = 8,640 Wh/day ≈ 8.6 kWh on a clear spring day

That same array in December at 2.2 sun hours and slightly cooler panels:

2,400 × 2.2 × 0.85 = 4,488 Wh/day ≈ 4.5 kWh — about half, and still normal

On the instantaneous side: a 400W panel in bright midday sun with a cool cell temperature should read roughly 300–360W; on a hot afternoon, 250–320W is common. Solar panel output calculator

If your actual production is within about 10–15% of this baseline, you don’t have a fault — you have weather, season, or a measurement-timing issue. Below that gap, work the diagnostic in order.

The diagnostic: cheapest check to deepest

Step 1: Visual inspection (free, 10 minutes)

Walk the array at the time of day output normally peaks.

• Soiling: look for an even gray film (pollen, dust) versus localized splatter (bird droppings, leaf tannins). Uneven soiling hurts more than a uniform film because shaded spots behave like partial shade. See what cleaning actually costs before renting gear.
• Shading: trees grow; vents, satellite dishes, and new construction cast new shadows. A shadow that crosses the array at 2pm but not 10am will hide if you inspect at the wrong time. How badly shade hurts depends on your wiring — see how shading affects panels and what bypass diodes do.
• Obvious damage: cracked glass, burned or melted connector plastic, chewed wiring, panels flush with snow.
• Monitoring trend: if you have panel-level or string-level monitoring, one weak string or module identifies the suspect before you touch a meter.

Step 2: String voltage and current math (multimeter, 20 minutes)

With a basic multimeter and your panels’ spec sheet, you can prove or clear the entire input side.

Open-circuit voltage (Voc): disconnect the string (or measure at the controller’s PV terminals with PV disconnected) and measure in decent sun.

• A nominal “12V” panel should read 18–22V in sun.
• A typical 60-cell residential panel reads within about 10% of its spec-sheet Voc (commonly 38–50V depending on model).
• Series strings add: ten 40V panels ≈ 400V. If your string reads half of what the math predicts, a panel or connection in the string is open or bypassed.

Short-circuit current (Isc): briefly measure through the meter’s amps jack (or use a DC clamp meter around one conductor — safer, no disconnection). A panel’s Isc should land within about 10% of spec. A reading far below spec with clean, unshaded glass points to a failing panel or a resistive connection.

Voltage at the controller: for battery-based systems, PV voltage should sit at least ~5V above battery charging voltage or an MPPT controller has nothing to track. A 12V panel (≈18–22V) feeding a 24V battery bank in parallel wiring can’t do that — that’s a design problem, not a fault.

Hot terminals, discoloration, or melted connector insulation at any point: stop and stop using the circuit — that’s a fire risk, not a performance problem.

Step 3: Controller and inverter logs (free, 15 minutes)

Input checks out? Look at the electronics’ side of the story.

• Error and derating codes: over-temperature derating on hot inverters, grid-voltage faults, and PV overvoltage on cold mornings are common and often self-clearing. Look up codes rather than guessing.
• Charging stage: bulk/absorption taper and float are normal. A full battery showing 2A in the afternoon is the system working correctly. Our MPPT not charging checklist covers the stage-by-stage detail.
• Settings: a lithium battery on a lead-acid profile (or vice versa) chronically undercharges. Verify battery type and charge voltages against the battery manufacturer’s sheet.
• Timestamps matter: output that drops every afternoon at the same time is shading; output that’s low all day on clear cool days is more likely soiling, wiring, or a hardware fault.

Step 4: Isolate panels (deepest check)

If a whole string underperforms, find the individual culprit.

1. With PV disconnected and the meter on DC volts, measure each panel’s Voc one at a time. One panel far below its neighbors is your suspect.
2. Measure Isc on the suspects. Low Voc and low Isc on clean glass usually means a failed cell group or a bad bypass diode.
3. On battery-based systems you can also disconnect panels one at a time (in series strings, removing a shaded or failed panel often lets the rest of the string recover) and watch charge current respond.

Replace or professionally repair the outlier; don’t keep running a panel that reads wildly off-spec.

Winter is not a fault

The single most common false alarm: “my system lost half its output.” If the drop began in late fall, do the seasonal math before touching anything.

• Shorter days and a lower sun angle cut daily production 40–60% in most of the US between July and December. A system that made 2,000 Wh/day in July can legitimately make 600–900 Wh/day in December at northern latitudes — same panels, same wiring, nothing broken.
• Peak sun hours drop more than people expect. Use your location’s monthly average (NREL’s PVWatts is the standard free reference) rather than guessing from summer numbers.
• Cold actually helps panel efficiency. Silicon produces more voltage when cold — a clear, cold January afternoon can beat a hot July one hour-for-hour. Cold hurts output through shorter days and snow, not cell physics.
• Snow is a full shutoff, not a derate. A few inches of cover reads as 0W and is completely normal; it clears without intervention on most tilts.

When it really is the controller or battery

You’ve confirmed good input, honest wiring, and sane settings, and output is still low. Now the suspicion shifts — but check these two pages before buying anything:

• MPPT charge controller not charging — PV voltage headroom, cold-weather Voc limits, battery profile settings, and error codes. Most “dead controller” diagnoses turn out to be one of those four.
• Solar battery not charging — float behavior, BMS low-temperature cutoffs (lithium packs below freezing refuse charge by design), and net-load problems where the system produces fine but loads eat the charge.

And if the array charges the battery fine but household AC output is low, the input side isn’t your problem at all — see inverter keeps shutting off.

Safety first

• No roof work in rain, ice, high wind, or on wet panels. No production figure is worth a fall. Clean from the ground with a pole and deionized water, or hire it out.
• DC arcs don’t self-extinguish the way AC arcs do. Never disconnect MC4 connectors or breakers while current is flowing — open the DC disconnect or breakers first, then unplug connectors.
• Cover the panel with opaque material when measuring Isc to reduce arc risk at the leads, and use a meter rated for your system’s DC voltage and current.
• Melted insulation, burned smell, or hot connectors = stop. That’s an immediate-call-a-professional situation.
• When in doubt on anything above your head or inside a combiner box, hand it to a licensed electrician. Solar maintenance has a safe DIY core — see the maintenance checklist for what’s reasonable to handle yourself.

Panel reads completely dead? (zero output)

A panel that reads 0 volts in full sun is a broken electrical path, not a mystery — and it is almost never the controller or the battery. The fault sits somewhere between the silicon and the controller terminals, and a blanket plus a multimeter pin it down in about 20 minutes. Work in this order: isolation test, Voc/Isc at the leads, continuity through the wiring, then a verdict.

Step 1: the two-minute isolation test

Do this first because it is free and it clears the controller side of the question.

1. On a clear day with the system producing, note the controller’s PV voltage and charge current.
2. Drop an opaque blanket or tarp over the suspect panel.
3. Wait 60–90 seconds and watch the display. Healthy response: PV voltage collapses toward 0 and charge current falls while covered; both recover within a few seconds of uncovering.
4. No reaction at all — the controller keeps showing the same numbers whether the panel is covered or not — combined with a 0V Voc reading at the panel’s own leads in full sun points to the panel or its wiring path, not the charge controller. (If the controller did react, the panel side is fine and the fault is downstream — see solar battery not charging.)

If the controller has no live voltage readout, do the same test with the meter on DC volts at the panel’s leads: covered ≈ 0–2V, uncovered jumps back to full Voc.

Step 2: meter basics — Voc and Isc at the panel, not through the wiring

Disconnect the panel’s leads at the controller end (or at the extension cable) and measure right at the panel’s MC4 pair, so a bad connection downstream can’t masquerade as a dead panel.

• Voc (open-circuit voltage): meter on DC volts, leads across + and −, panel in full sun. A 12V-nominal 100W panel should read roughly 18–23V — lab spec sheets commonly list 21–24V at STC (25°C, 1000 W/m²), and a cool bright afternoon runs higher, not lower. Expect within ~10% of the number printed on the label.
• Isc (short-circuit current): switch the meter to DC amps (leads in the A jack) or use a DC clamp meter around one conductor. Touch across + and − for a few seconds — it’s a short by design. A 12V-nominal 100W panel should read roughly 5–6.5A and within ~10% of its label. Quick sanity math: 100W ÷ ~18V Vmp ≈ 5.5A at max power, and Isc runs a bit higher than that.
• Both readings fall with partial sun. Voc especially collapses in shade, clouds, or early morning — never condemn a panel measured in anything but full sun on clean glass. (Cover the glass while connecting the leads to reduce arc risk, then uncover to measure — same practice as the safety notes below.)

Step 3: bypass diode failure — the “output cut in half” symptom

Most 12V-class panels have three (some four) bypass diodes, each protecting one section of cells. When a diode fails shorted — the common failure after years of shading and thermal cycling — the section it protects stays out of circuit permanently:

• The panel still produces — the diode gives current a path around the dead section — but voltage and power drop by that section’s share: commonly a third to a half of the panel’s healthy Voc, which is why a dead diode section reads like “output cut in half.”
• In a series string this shows up as one panel a clear step below its neighbors’ Voc while the string’s current looks normal — the diode carries current, so nothing downstream trips.
• You can confirm it two ways: shade one section of an otherwise healthy panel and Voc drops by only that section’s expected share (normal); a dead diode keeps that same drop in full direct sun with nothing shaded. Or, in an accessible junction box, use the meter’s diode mode — a shorted diode beeps in both directions; a good one conducts one way only.
• An open bypass diode (rarer) hides in full sun, but under partial shade it forces the shaded cell group into reverse bias — that heats the group and is exactly the scenario behind localized burn marks and hot spots.

Step 4: connector and inline-fuse continuity

With the panel disconnected and covered, meter on ohms, check end to end: panel MC4 pair → extension cable → inline fuse → controller’s PV terminals. Expect near-0Ω (beep). An open reading means a loose or corroded MC4 (tug and look — cheap pairs are a classic), a chewed conductor, or a blown inline fuse — pull the fuse and check it out of circuit, because fuses can fail open under load and read fine in circuit.

When it’s really dead — and the honest salvage note

Call the panel dead when all of the above is done and cleared: full sun, clean glass, good continuity, no diode short — and it still reads 0.0V at its own leads. Add physical tells and the meter doesn’t even matter: cracked glass (water gets inside and it will fail eventually anyway), delamination (bubbles or haze under the glass), or burn marks under the glass — those are replace, do not operate, cases regardless of readings; a burned panel can present an electrical-safety and fire risk.

Salvage honestly: cracked-glass and burned laminates have no safe DIY repair and no salvage value as panels — route them to a PV recycler; the aluminum frame is the only part worth recovering. A shorted bypass diode in an accessible junction box can sometimes be replaced (desolder the diode, solder in a same-rating spec, re-seal with di-electric compound) — but only if the laminate is sound and you are confident soldering; otherwise the “repair” invites moisture ingress and a shorter-life panel. If the panel is salvage-economically on the fence — no glass damage, just a repairable diode — a 50%-output panel still charges better than a garbage can, but it is not worth re-installing in a prime array spot.

For the controller side of the same symptom set, work the MPPT not charging checklist; for expected output numbers from a healthy panel, the solar panel output calculator gives the baseline any diagnosis should be compared against.

FAQ

Next logical reads

Solar panel output calculator (derating factors) How shading affects solar panels Solar panel cleaning cost (DIY vs pro) Solar maintenance checklist MPPT charge controller not charging Solar battery not charging

Related guides:

• MPPT Charge Controller Not Charging: Troubleshooting Checklist (PV Voltage, Settings)
• Solar Battery Not Charging: Troubleshooting Checklist (MPPT, Wiring, Loads)
• DIY Small Wind Turbine for Battery Charging (Wiring + Diversion Load Control) Test a solar panel with a multimeter