Why Solar Garden Lights Go Dim After One Season

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

They mostly have not stopped working. They have reached the number of charge and discharge cycles they were specified for, which in a garden light is a surprisingly small number, and they got there on schedule.

The giveaway is the order the symptoms arrive in. A solar light that has been flooded or whose switch has corroded goes dark all at once. A solar light that has simply aged goes dim first, for weeks, and the dimness is the diagnosis.

Dim is a battery symptom, and it is a voltage symptom

An LED is not a bulb. A filament glows in proportion to the power you hand it and will glow faintly on half a volt. An LED passes almost no current until the voltage across it reaches its threshold, then its current rises very steeply. So the brightness of a solar light is set by the voltage its cell can hold up while current is being pulled out of it, not by how much charge is left in the cell.

Two things happen to a small rechargeable cell as it is cycled. It loses capacity, which shortens the night. And its internal resistance rises, which means that every milliamp the LED draws costs more voltage inside the cell than it used to. The second effect is the one people misread as a fault, because a cell can still be holding plenty of charge and yet be unable to deliver it at a voltage the LED will answer to.

The cycle arithmetic, which is the whole answer

A garden light does something a battery engineer would wince at. It discharges its cell as far as the circuit allows, every single night, and then fills it again the next day. That is 365 deep cycles a year.

Cycle life for a small nickel metal hydride cell is quoted in the hundreds at that depth, not the thousands, and the depth is the reason. Run the same cell to half empty and you get several times as many cycles out of it; run it flat and the count collapses. So one to two seasons of full-night running is not a defect, it is the design, and it follows from arithmetic that was knowable before anyone bought anything.

Our own guide to how solar garden lights work says the battery gradually holds less charge after a few years. A few years is too generous for a light that genuinely runs all night, every night, and the correction matters because it changes what you do about it: at a few years you shrug and buy new lights, at one to two seasons you buy a cell for pennies.

Now read the runtime figures again, because they are brightness figures

Here is the part that reframes the whole category. The energy a solar light can spend in a night cannot exceed what its panel collected that day, and the panel's ceiling is set by its area and by the sun, both of which you can look up.

Full noon sunlight delivers roughly 1,000 watts per square metre. A garden stake panel about 1.6 inches square is 16.5 square centimetres, so it intercepts about 1.65 watts of light, and a cheap cell converting at fifteen percent can make about a quarter of a watt at its very best. Over a good day worth four and a half full sun hours, that is about 1.1 watt hours. That is the entire budget.

So take a set of Twinkle solar garden lights in an 8 pack, which publish a 6 to 8 hour charge giving a stated 14 to 16 hours of light from fifteen flickering LEDs behind a 3D laser film, in a waterproof ABS and acrylic body. Spread 1.1 watt hours over 15 hours and you get about 73 milliwatts of LED power. Share that between fifteen LEDs and each one is running at about five milliwatts, which is roughly an eighth of what an ordinary indicator LED on an appliance draws. At the efficiency these run at, the whole stake is putting out something in the region of five to seven lumens. A household candle is about twelve.

That is not a criticism of the product, which says so itself: its own record lists decorative ambient light only, not bright task or security lighting, and that the charge and runtime figures are the maker's best case. It is a way of reading the specification. A long published runtime on a solar light is a statement about how little the LED is drawing, which is to say about how dim it is, and the two numbers cannot both be large. Our review of the Twinkle solar garden lights has the rest.

The same reading makes a striking claim make sense. A set of AGPTEK solar driveway lights with three colour modes and an IP67 alloy body publishes 30 hours of runtime after a 5 hour solar charge, six LEDs, five modes, a 2 ton pressure rating, and a detachable stake. Thirty hours of light from five hours of charging means the LEDs must be spending power at roughly a sixth of the rate the panel was gathering it. Run the same area arithmetic on a panel about two inches square and you get perhaps 1.2 watt hours a day, which over 30 hours is about 40 milliwatts across six LEDs. And then notice what the record says is missing: brightness in lumens is not listed. The 30 hour figure and the absent lumen figure are the same fact written two different ways. Details in our review of the AGPTEK driveway lights, and the category view is in our guide to choosing solar driveway lights that last.

A linseyCove outdoor solar clip light with IP65 and three modes is the most internally consistent of the three, publishing a 6 to 8 hour charge for 8 to 12 hours of light, a polycrystalline panel, coverage up to 30 square metres and a dusk to dawn sensor. Eight to twelve hours is a modest claim and the coverage figure is at least stated, which makes the pair of numbers checkable against each other rather than only against the night. Its own record carries the caveat that matters most for everything below: winter daylight will reduce charging. See our review of the linseyCove clip light and our guide to choosing and placing solar clip lights.

Where the energy balance stops closing, and the listing tells you

The budget above is per square inch of panel, so the moment a product's appetite grows faster than its panel, the design has to find power somewhere else. Watch for a second input, because it is an admission.

A run of SUNTHIN 100ft solar string lights with USB recharging publishes 100 feet of shatterproof bulbs, a dusk to dawn sensor, an IP65 build, and a bundled USB cable so a power bank can keep the string lit on cloudy days. That last feature is the honest part of the listing. A hundred feet of bulbs asks for far more than a string light panel gathers, so the USB port exists because the arithmetic does not close, and its own record is frank that neither brightness nor runtime figures are provided at all. More in our review of the SUNTHIN solar string lights.

The extreme case is worth doing in full because it shows how far the gap can go. A RadiantFlux rechargeable solar camping lantern with a 6600mAh cell publishes a stated 4000 lumens, three modes, IP65, a magnetic back, a quarter twenty thread, a USB output for charging a phone, and both solar and Type-C input. A 6600 milliamp hour lithium cell at 3.7 volts holds about 24 watt hours. A panel the size of the top of a lantern, call it three inches square, gathers something like 4 watt hours on a good day. Filling that cell from empty by sunlight alone would take about six good days.

And the output side settles it. Even if the 4000 lumen figure were conservative, getting anywhere near it needs tens of watts, so a full cell is measured in tens of minutes on high and a day of sunshine buys single-digit minutes. The Type-C port is not a convenience, it is the power supply, and the panel is a trickle charger. Its own record flags exactly the two things that follow: no wall adapter is included, and the per-mode runtimes are not specified. Our review of the RadiantFlux solar lantern sets out what is and is not published. None of this makes it a bad lantern. It makes it a rechargeable lantern with a solar panel on it, which is a different product from a solar lantern, and the distinction is worth having before you rely on one.

The honest limit

Every assumption above is marked because none of these listings publishes a panel rating or a cell chemistry, which is the real gap in the category. Panel areas here are estimated from the physical size of the products, cell efficiency is taken at fifteen percent, and a good day is taken as four and a half full sun hours. Use your own figures and the conclusions move, but they do not reverse, because the panel area on a garden light is a couple of square inches and that ceiling is not negotiable.

The fix is almost free, and it is the cell. In most stake and path lights it is a single standard rechargeable AA or AAA sitting under a screwed cover, and swapping it restores the voltage, which restores the brightness. Two things to get right while you are in there. Buy the low self discharge type, the ones sold pre-charged, because an ordinary cell loses something like fifteen to twenty percent of its charge a month just sitting there while the better sort loses roughly that much in a year, and in a shaded winter garden that difference is most of the problem. And do not buy the biggest capacity on the shelf: a cell three times the size takes three times as long to fill from a panel that has not changed, so it will simply never reach full and you will have bought a cell that lives permanently in the part of its range that ages fastest.

Then stop asking them to go flat every night, because depth of discharge is the single biggest lever on cycle life. Switching a decorative light off for part of the week, or running a mode that quits before dawn rather than one that limps to it, is free and it roughly multiplies the seasons you get.

And keep the two separate questions separate. How long a light runs on a full charge is a different matter from how long the product lasts, and we did the first one properly in do solar lights stay on all night, which compares the published runtime to the actual length of the night month by month. This page is about the second. Together they explain almost every complaint in the category without anything having broken.