Deck Post Cap Lights: Hardwired vs. Rechargeable Lithium—Which Lasts Longer on a Salt-Air Deck in Maine?
You install post cap lights on your cedar deck in Kennebunkport thinking, “This’ll look clean, modern, safe.” Six months later, one fixture flickers like a dying firefly. Nine months in, three are dead. Twelve months? Half the set’s gone dark—or worse, green with corrosion at the base. You didn’t buy cheap junk. You read reviews. You paid extra for “marine-grade.” Still, salt air won, and it did it quietly, relentlessly.
Here’s the mistake most people make: they treat deck post cap lighting like indoor lighting—with indoor assumptions. They pick based on lumens or aesthetics, then bolt it down and forget it. But on a Maine coast deck, you’re not fighting time. You’re fighting electrochemical warfare: sodium chloride dissolving copper traces, freezing lithium cells cracking internal dendrites, 12V DC lines bleeding voltage over cedar-framed runs while snowmelt pools under railings.
I tracked two identical sets of post cap lights—same optics, same IP67 rating, same 300-lumen output—on my own Kennebunkport cedar deck (32’ × 18’, 8 posts, southern exposure, full salt spray from the Saco River estuary) for exactly 12 months. One set hardwired to a transformer-fed 12V DC low-voltage system. The other: rechargeable lithium-ion caps with integrated 2,200 mAh cells and solar charging panels. No batteries swapped. No firmware updates. Just real-world abuse: -14°F nights, 90% RH fog banks rolling in off the Atlantic, and visible salt crust on railings every March.
Let’s cut the marketing fluff. There is no “best” option. There’s only *which failure mode you can tolerate*.
The Hardwired Myth: “Just Run Better Wire and You’re Golden”
That’s what the landscape lighting installer told me. “Use 10 AWG stranded tinned-copper wire, run it in PVC conduit above the joists, and terminate with marine-grade Wago lever-nuts. Done.” Sounds solid—until you see the voltage drop chart.
I measured voltage at the transformer (12.12V), then at each post—every 4 feet—using a calibrated Fluke 87V. Here’s what the 30-foot run actually delivered:
| Distance from Transformer | Measured Voltage (DC) | Lumen Output (Relative %) | Observed Flicker/Instability |
|---|---|---|---|
| 0 ft (at transformer) | 12.12 V | 100% | None |
| 8 ft | 11.94 V | 98% | None |
| 16 ft | 11.71 V | 92% | Barely perceptible at dusk |
| 24 ft | 11.43 V | 83% | Faint pulsing during wind gusts |
| 30 ft (last post) | 11.27 V | 76% | Steady 2Hz dimming; visible at night |
This isn’t theoretical. At 11.27V, the driver ICs in these particular caps—designed for stable 12.0–12.6V input—begin throttling current to protect themselves. That’s why lumen maintenance dropped 24% at the far end. Not because the LEDs aged. Because the power supply couldn’t hold its own.
And that’s *before* corrosion hit.
By Month 5, I found white crystalline deposits (sodium chloride + copper hydroxide) forming where the wire entered the post cap housing—even though the connector was rated IP67 and sealed with Loctite SI 598. Salt doesn’t need a breach. It wicks. It creeps. It condenses in micro-gaps overnight when temps swing from 28°F to 41°F. By Month 9, two of the five downstream caps showed intermittent open-circuit faults. Multimeter confirmed: continuity lost at the PCB trace near the terminal block—not the wire, not the nut, but the copper itself, eaten thin.
I replaced those two with new caps—and resealed with Dow Corning 3145 (a true marine-grade conformal coating). They lasted until Month 11. Then one failed with a thermal runaway smell. Turns out, the corroded trace had increased resistance just enough to generate localized heat at 11.3V. Not enough to trip protection. Enough to cook the epoxy encapsulant and delaminate the LED substrate.
Hardwired systems don’t “last longer.” They fail slower—but more catastrophically. And they demand ongoing vigilance: annual voltage mapping, biannual terminal inspection, quarterly cleaning with distilled vinegar + baking soda paste (not pressure washers—those force salt deeper).
The Rechargeable Lithium Pitch: “Set It and Forget It”
“Solar-powered! No wires! Zero maintenance!” reads the box. What it doesn’t say: “Your battery will lose 32% of its capacity after one winter if left fully charged at -10°F.” Or: “That ‘self-cleaning’ solar panel? Salt crust reduces its output by 68% in February.”
My lithium set used monocrystalline 2.5W panels (22% efficiency), 2,200 mAh Li-ion cells, and charge management ICs claiming -20°C to 60°C operation. On paper: fine. In Kennebunkport January? Brutal.
Here’s the cycle degradation curve I logged monthly using a benchtop battery analyzer (not manufacturer specs—actual discharge curves at 300mA load):
- Month 1: Full 2,200 mAh capacity. 6.5 hrs runtime at full brightness (300 lm), 11.2 hrs at dim (120 lm).
- Month 3 (January): 1,940 mAh. Noticeable dimming after 4 hrs. First “low-battery blink” at 8:12 PM (sunset 4:28 PM).
- Month 6 (April): 1,680 mAh. Solar panel output measured at 0.8W avg. in overcast coastal spring—down from 2.1W in August. Runtime now 3.2 hrs full-bright.
- Month 9 (July): 1,520 mAh. Panel cleaned weekly—still 15% less output than new due to micro-scratches from salt abrasion. Thermal throttling kicks in above 32°C ambient (common on cedar in full sun).
- Month 12 (December): 1,310 mAh. One cap completely dead—cell voltage collapsed to 1.8V and won’t accept charge. Two others require manual reset via hidden button sequence (buried under silicone seal) to wake from deep sleep.
This falls flat because the chemistry wasn’t matched to the environment. Lithium cobalt oxide (LiCoO₂) cells—the kind used here—have terrible low-temp charge acceptance. Below 0°C, the anode kinetics slow so much that lithium plating occurs instead of intercalation. That’s irreversible capacity loss. And no solar panel, no matter how “high-efficiency,” generates meaningful current when covered in salt haze and angled at 43° north latitude in December.
Also: the “rechargeable” claim hides a truth—these aren’t user-replaceable. The cells are spot-welded. No datasheet lists cycle life at -15°C. No UL file covers thermal runaway risk in enclosed cedar post cavities. When one failed, I cracked it open. Found dried-out electrolyte residue and a swollen cell pressing against the lens gasket. That gasket failed at Month 10. Salt got in. LED array shorted.
So yes—they’re wire-free. But “maintenance-free”? No. They demand seasonal recalibration: tilt-angle adjustment for winter sun, biweekly panel scrubbing with 50/50 isopropyl alcohol/water (vinegar etches anti-reflective coatings), and firmware resets when cold-soak confuses the BMS.
What Actually Survived 12 Months—And Why
Neither setup “won.” But one held up with far less intervention.
The hardwired set had 3 of 8 caps fully functional at Month 12. All three were within 12 feet of the transformer. The rest required either replacement or voltage-boost mods (I added a local 12V boost converter at Post 6—worked, but looked janky inside the cedar post).
The lithium set had 4 of 8 caps still lighting nightly—but only at reduced brightness (120 lm), only after manual wake-up, and only with panel cleaning done religiously. Two were bricked. One lit erratically—turning on at 2:17 AM, off at 3:04 AM—because its light sensor was fogged internally.
But here’s what surprised me: lumen maintenance across *both* groups was nearly identical at the working units. After 12 months, surviving LEDs retained 89–91% of initial output. Why? Because LED degradation is driven by junction temperature and drive-current stability—not by whether power comes from wire or battery. Both systems kept drive current tight within spec. Salt didn’t kill the LEDs. It killed the delivery system around them.
The real durability differentiator wasn’t the power source. It was the housing construction and sealing method.
The caps that lasted longest—across both groups—shared three traits:
- Over-molded silicone gaskets (not foam tape or rubber O-rings), compressed to 30% deflection at assembly.
- No exposed metal fasteners—stainless steel 316 screws only, with nickel-plated brass inserts embedded in the housing (no threads cut into aluminum).
- Conformal-coated PCBs, including edge connectors—applied *after* final assembly, not just on bare boards.
The ones that failed first? Used zinc-plated steel mounting brackets. Had vent holes “for thermal relief” (a myth—heat escapes via conduction, not convection, in these tiny enclosures). Or relied on a single bead of RTV silicone that shrank and cracked after UV/salt exposure.
So—Which Should You Choose?
If you want predictable, serviceable, long-term performance—and you’re willing to invest upfront in infrastructure—go hardwired. But do it right:
- Use a regulated 12V DC switching transformer (not magnetic tap-type), with remote voltage sense leads run back to the last fixture.
- Size wire for actual load + 25% margin—not catalog tables. For 30 feet and 8×3W loads, I upgraded to 8 AWG tinned-copper. Dropped voltage drop to 3.1% at the far end.
- Install a weatherproof junction box at each post with dielectric grease on every termination. Not optional. Salt finds the gap.
- Accept that you’ll replace 2–3 caps every 24–30 months. Budget for it.
If you want simplicity, minimal visual impact, and you’ll commit to hands-on seasonal care—go rechargeable. But skip the “solar-only” models. Get hybrid units: lithium with USB-C external charging ports *and* solar topping-off. That way, when February rolls in and your panels produce 0.3W, you plug in a $20 Anker PowerCore for 2 hours once a month and keep the cells at 40–60% SOC—where lithium degrades slowest.
I think the hybrid approach wins for most coastal homeowners—not because it’s perfect, but because it trades one kind of labor (wire pulling, voltage mapping) for another (monthly top-ups, panel cleaning) that’s less invasive and more forgiving.
One last note: cedar matters. Its natural tannins accelerate copper corrosion. I switched all hardwired terminations to brass-to-brass connections after Month 6. No more copper wire directly contacting cedar sap channels. That alone added 4 months to the life of the downstream caps.
There’s no magic bullet. Only trade-offs measured in volts, milliamps, sodium ions, and patience. Your deck isn’t a patio in Denver. It’s a laboratory for material failure. Respect the salt. Measure the voltage. Clean the panels. And never, ever trust the IP rating on the box.