Hacking Color And Battery Life On A Solar Powered Garden Light

Give your garden lights a little bit more life by changing out the LEDs. In the process, you can also choose a different color if you like. It really is as simple as cracking it open and replacing the stock LED, but a bit of a change may also prolong the stored charge.

These garden lights consist of a small solar cell and an LED, both controlled by a small PCB powered from a NiMH rechargeable battery. Sure, you could try to put in a battery that has more potential, but if you replace the single LED with two of them in series, it drops the current consumption and increase the battery life. Just make sure to use super bright LEDs and the intensity change will not be all that noticeable. That comes partly from the fact that our eyes don’t detect intensity changes as well when a light is already very bright. And since it’s two LEDs, mixing colors is an option as we see above.

[Thanks Ken]

15 thoughts on “Hacking Color And Battery Life On A Solar Powered Garden Light”

1. Cliff says:

This is a fun project, but the second paragraph is nonsense. Putting two LEDs in series does nothing for the amount of current flowing through each (since current is conserved in series). The voltage drop across each diode will be lower, since the sum of voltage drops around a loop is zero.

Putting two LEDs in -parallel- will result in less current flowing through each (since the current is divided between both branches), but the total amount of current remains the same.

1. Gorloth says:

Actually putting the LEDs in series will result in lower current, the LEDs are being supplied a constant voltage (technically the voltage off the battery will drop slightly as it discharges), if you split that voltage between two LEDs then each one will have ~half that voltage across it resulting in lower current draw per LED (less voltage across an LED results in lower current draw) putting the LEDs in parallel results in more current draw as each LED has the same voltage across it as a single LED did and thus has the same current.

You can think of the LEDs as resistors (not entirely accurate as LEDs has a very non linear I/V curve) putting two in series means more resistance, thus lower current, two in parallel means lower resistance, thus more current

2. Greenaum says:

Cliff, are you sure? Cos that’s the opposite of how I thought simple electrics worked.

What quite likely IS a problem, is that these lamps use a DC-DC boost converter, to get LED-driving voltage from a 1.2v cell. Putting 2 LEDs in series will demand a higher voltage from the converter, and since energy doesn’t come from nowhere, the converter will use more current.

Unless the guy has measurements that say otherwise. And even then, adding a single LED to a commercial product isn’t the most amazing hack ever posted here.

1. NewCommentor1283 says:

this guy is closer(est) to the truth

the boost converter used in totally UNREGULATED.
the boost converter in solar garden lights ect is actually more of a flyback then a booster/converter

… which means;
the lower voltage the led the more current will flow through LED and more will be drawn from battery. the ones ive seen do NOT draw more current from higher LED voltage, that would be in a regulated converter, but in this one it is funky math that will show you the way.

look up “flyback” converter

1. NewCommentor1283 says:

i got around 7 or 8 volts peak with NO load.
around 6 volts with ~1ma draw and the usual 5-10ma with ~3v LED

my unit with 2.2v GREEN led hasnt bitten the dust yet, so im assuming they wont go over 20ma with a 2.2v led, then again who says the original was running the white (3.5v) led at full brightness?

PS: after operating one with no load for an hour the frequency/dutycycle started changing drastacly, maybe oscillating?, so the parts(transistor/fet) get slowly destroyed from voltages over 6v? hmm

3. echodelta says:

Solar garden lights are very bad for the Earth our garden. I watch a set get set out, then watch the duds appear one after the other as batteries with cadmium corrode and fail freely. Squirrels disassemble them leaving things at the mercy of rain. Better to use low voltage led wired lighting. What is insulting is when is when the owner places some in the shade, and then throws out them all when they aren’t working. Like a battery necessary in something that is plugged in, battery operated stationary things on an electrified property don’t make sense.

1. NewCommentor1283 says:

usually not the lamp’s fault, it’s the user.

they are NOT winter compatible OR shade compatible,
people seem to ignore this all the time ugh
oh and you HAVE to sand battery contacts/battery EVERY FRICKING SPRING!!! duhhh!!!!!!! )(&(^&)*_*&)(_)

… or half them wont work eventhough they contain perfectly working parts.

lol reminds me of the catchphrase:
“IT’s THE REST OF THE WORLD THATS CRAZY! NOT ME!” lolz

PPS: they need RTV compound on the LED’s pins,,, rust goes right into the LED’s insides…

1. NewCommentor1283 says:

do i need to add in they should not be stored underwater… funny story hehehehe

2. Tony says:

@echodelta is pointing out that if they were made properly then you wouldn’t need to do all that.

Still, what do you expect for \$2? (Well, a pile of useful parts in my case, and most others here too.)

BTW, you can get RGB colour changing ones…

4. Mysterio says:

What Cliff says would kinda make sense if the was being driven by a constant current source and there was no current limiting resistor. It depends really… the guy doesn’t mention any current limiting resistor on the board, and we don’t really know what’s going on with that chip. *IF* you make some wild guesses on the numbers and model things as being a 12vdc supply driving an LED with a 3.3V forward voltage at a nominal 30mA forward current, and calculate the optimal current limiting resistor you come up with 290 Ohms. (R = V/I = (12 – 3.3)/(30 x 10-3) = 290). Now, if you just drop a second identical LED in series but don’t alter the current limiting resistor, and make the assumption that both LEDs are operating at their full forward voltage then you can calculate the current as being I = V/R = (12 – 6.6)/290 = 18.6 mA. So, yeah obviously the current is reduced. Because of that the diodes wouldn’t truly be operating at the full 3.3V forward voltage drop, things would kind of balance out there.

Current is not magically conserved in series unless it’s a constant current source. Adding or removing any components is a change in the load.

1. NewCommentor1283 says:

you are correct about DC BATTERY POWER!

we are NOT dealing with this.

we are dealing with a flyback, it has unique and ANNOYING math involved, its better to measure instead of spend all day just to come up with bad math on this one, in my opingion.

PS: i think some of them, ALSO do PWM to REPRESENT andor reinforce the effect the battery getting low… make it last an extra hour

2. Cliff says:

I think the LED lights you guys are working with are perhaps a bit higher-end than the ones I’ve dissected. I made some assumptions here — constant source, same resistance in each LED, nothing fancy. I figured I’d cut it off before it got all long-winded, oops. :-P

5. Use an HV850 and two opposed strings of series white LEDs salvaged from a broken laptop’s LCD panel to get a very low powered lamp indeed.
These beasties have the advantage that the frequency can be microcontroller set and also draw under 18mA fully loaded with a series capacitor.

These also run from down to 3.1V so a lithium ion can be used with a MAX1555 as the charger.

6. Bubba Gump says:

It is all dependent on the quality of the light. A 3 to 4 dollar light at wally world does not have the quality of a light that retails for 10 to 20 dollars.

The boost circuits typically take a 1.5 volt cell and boost it to ~ 3.2 volts for a white LED. If there are multiple led’s typically they put them in series and boost the voltage and keep the current at 20mA or so on many of the units I have personally torn down.

On many the boost IC is COB (Chip On Board) but the pc boards are low quality (yes, I have seen COB of CEM-1 and CEM-3 boards…shudder…)

The human eye does not perceive light in a linear fashion. 1 lumen of a white led is perceived less bright than a green led also putting out 1 lumen. Our eyes perceive peak light at 490 to 560 nanometers.

An integrating sphere will show cold hard numbers but your brain and the rods and cones in your eyes will tell you otherwise.

There have been all sorts of studies and TONS of regulations regarding street lighting. It is dry reading but quite interesting to read, however, you probably will have to identify with Sheldon Cooper and say BAZINGA to want to read it.

There are colleges out there, believe it or not, you can get a master’s degree in lighting science.

7. I did this last year for Ligtning in a Bottle with a crate of dollar store solar lights; many of these lamps use a tiny tiny battery and a joule theif like circuit to power the stock white LED, I noted that the lower-power LEDs like yellow through red die much faster. UV, blue (and pink and white), cyan and green LEDs all last the night.

This site uses Akismet to reduce spam. Learn how your comment data is processed.