Outdoor Lantern Solar Solution needed
 

Greetings,

I've been all over Amazon and AliExpress looking for a type of product that boils down to an LED bulb connected to a solar panel. Almost all options are a dim bulb with a tiny solar panel.

I don't understand is why someone doesn't offer way brighter bulbs with a way bigger solar panel. (Is it a battery storage issue?)

So I visit here under the presumption I may need to hire an electrician to 'make' the product I'm seeking. That is: I take a lantern and bulb combo I already use and juice it (somehow) with a larger solar panel.

I'm going to include links to my bulb and lantern. I must stick with these because I'm in a condo complex using only this fixture. (What I'm doing is adding a lantern where no electricity ever was set up.)

This is not spam. It's just so that anyone who has spec questions can simply go to the links.

The lantern --

https://www.homedepot.com/p/Acclaim-...W-FR/204204459

This Dusk Till Dawner bulb (7 Watts) -- (edited, explained way below)

https://www.amazon.com/Sensor-Lights-Lighting-Automatic-Outdoor/dp/B07KFML3CH/ref=pd_sim_60_3?_encoding=UTF8&pd_rd_i=B076CN463G& pd_rd_r=708b4f24-1065-11e9-a4fb-b1d8f98b2ea8&pd_rd_w=WnFrk&pd_rd_wg=jXLfp&pf_rd_p= 18bb0b78-4200-49b9-ac91-f141d61a1780&pf_rd_r=XP9K00YH4MRRENJKQ6X2&refRID=X P9K00YH4MRRENJKQ6X2&th=1

So I'm looking for a way to set up a large enough solar panel/battery/convertor situation to power this.

In this picture I show where I'd like to run the exposed wiring.

https://photos.app.goo.gl/MJ378Vd5SPSvj2Sx7

The panel would run to a California rooftop typically bathed in light.

So is this possible? And reasonably affordable?

Any help would be greatly appreciated.

I would think the battery and charger are the hardest part?

Are you going with an all DC set up?

If you're asking would I use the wires that come out of the lantern, yes. Since I'm (frankly) unaware of alternatives. Frankly = clueless on electricity.

My concern about a battery is that in summer it can get HOT where I am. I can come up with schemes of where to hide it but my concern is if a battery can simply spend weeks at a time over 100 degrees Fahrenheit.

When you say 'charger' do you mean a mechanism that instructs the solar to be saved to the panel? Or from the panel to the lantern/bulb? Or both?

Thanks for your response.

(By the by, I'm originally from Brockton.)

Hi,

The solar panel produces DC - direct current - electricity. The battery also is DC. The bulb could be DC, but many are AC - alternating current; which is what your house is. The battery charger regulates how much power goes into the battery. The wiring for DC is quite different than AC - it has to be larger gauge to reduce losses, the longer the wires are.

It looks like the bulbs you have chosen are AC. So, you need an inverter to get AC from the DC battery.

Other pertinent questions. Is it being used for security? How many hours per night does it need to be on(battery size)? How much space do you have for a solar panel before the condo people get upset?

JJ

Great questions.

The security aspect is to augment existing lighting which is nearby. And so if there's a few days of the year the light doesn't really come on, that's okay.

I'd like the bulb to work all night but a definite minimum of 6 hours. This is a rooftop and so a panel size restriction in unnecessary. I don't know if a panel that's like 4 by 2 feet (or whatever) will cover this.

Is there a panel kit on Amazon that fits this bill?

Thanks for the Solar Schooling. :)

Is this project do-able?

Typical solar panels are 200-300 watts. The light you linked to is 5W, and depending on where you are, and how much sun you get on average - it should be very doable. But, it may or may not be "worth" doing.

It depends on the cost - if you can find a small (say 20-100 watt panel), for under $50, and a workable battery system for about the same - then would that be worth it to you?

A 5 watt bulb is a very small load. With a motion detector built into the bulb, it may only operate a few minutes a day - or maybe it won't run at all?

The expensive parts are the solar panel, the battery, it's charger, and the inverter - and the labor to install it all.

The good news is that the labor will be cheap. Under $200 for the rest would be acceptable. I've got California sun and no trees in the way and rarely clouds.

But that light isn't a motion detector -- it's meant to come on each night for as long as the battery can handle it. Minimum six hours.

The light you linked to is a motion sensor light. Neil has pretty much got you lined out. That bulb runs on AC. As he said you would need a very small inverter to turn the DC from the batteries into usable AC for the bulb. There are a number of other lamp solutions out there that do exactly what you want but not with that housing. If you must use that housing you will need the solar panel, a battery pack, the battery charge controller, an inverter and probably a small enclosure for the electronic items. You might try to find a direct DC LED bulb that will fit in the housing, remove the AC socket and find a way to mount the DC bulb. That would eliminate the need for an inverter.

Something like this could work.

https://www.amazon.com/GLW-Waterproof-Outdoor-Security-Equivalent/dp/B00DI3N0QM

JJ

You're right! Amazon changed the link to a different product. I went to my order of the previous bulb, clicked the link, and didn't notice they had changed the dusk till dawner to a motion sensor.

The good news for me is there's a new 7 watt Dusk till Dawner closer to my needs. It's here --

https://www.amazon.com/Sensor-Lights-Lighting-Automatic-Outdoor/dp/B07KFML3CH/ref=pd_sim_60_3?_encoding=UTF8&pd_rd_i=B076CN463G& pd_rd_r=708b4f24-1065-11e9-a4fb-b1d8f98b2ea8&pd_rd_w=WnFrk&pd_rd_wg=jXLfp&pf_rd_p= 18bb0b78-4200-49b9-ac91-f141d61a1780&pf_rd_r=XP9K00YH4MRRENJKQ6X2&refRID=X P9K00YH4MRRENJKQ6X2&th=1

So this conversation has changed slightly in that I'll be using a 7 Watt Dusk till Dawn light instead of 5 watt.

Hi Neil,

So I goofed you guys up with my bulb link. (Amazon had changed the product at that link and I hadn't noticed.) I altered the link to a 7 Watt Dusk till Dawner which keeps most of your science math similar.

So I'm getting ready to shop for parts but must ask re-ask a basic question -- do the battery, charger, and inverter require a cool space? Or are they designed to sit in hot California sun? If not --

-- would an enclosure help if it was sitting in that same sun? That is: am I water protecting the battery and charger and inverter or thermal protecting them too?

Here is the actual location and scenario I'm contemplating.

https://photos.app.goo.gl/tqTKWRXrDgYZq9N46

So that lantern would have a 7 watt dusk till dawn with almost the same lumens of the real lights throughout our complex. The red lines indicating wiring, which would likely be inside of think pipes. The battery/charger/inverter would be 'boxed' in plain site.

If this sounds about right is there a kit you recommend? I've been looking and it's kind of shocking that batteries for this purpose do not exist with inverters and chargers attached.

Yes, the box is more for weather/rain protection than heat. The best thing you can do for heat(or cold) is to run the wires from the solar panel into the house and connect all your electronics there. If that is not feasible then place some sort of cover above the box to keep it out of direct sunlight.
I don't know if there is a direct kit that would work for you. But if you are using a 7 watt bulb lets say for 10 hours a night then that would be a minimum of 70 watt hours per day. I would increase the battery capacity to at least 200 watt hours. You could probably get by with using two 12 volt 10 amp hour AGM batteries wired in parallel. You could use a 50 watt solar panel for charging. You have to be sure the panel is located so that it can get full sun for at least four hours per day even in the winter. JJ

If you do not want to know / learn solar
 

If you just want it usable system and do not want to know or learn all the ins and outs get a harborfreight.com/100-watt-solar-panel-kit-63585.html you will need to use both lights in my opinion and can get other DC lights latter if that is not enough.. wait till it is on sale and sign up for the coupons it will drop the price. Get one of there batteries if you are close to store but really a deep cycle will do you better and store more. The better the battery the more the cost. For what you want there battery should do the trick with out the hassle and be easy to replace. I used this set up to learn from and have a much more robust set up now. Any RV/camp trailer place sells 12 volt DC lights and bulbs if needed for more light.
It is not a high end system. Some people call it junk, and it is not that either. It is just proprietary connected system. Does not have a lot of ways to upgrade with out buying more of there equipment. At some point that maxes out what can be done. Hope that helps.

Just what I did
 

I ran the wires out side A shed and berried "battery box" in the ground with gravel under it and a piece of pvc for gas/pressure release along with the charge controller it was cooler in the summer and stayed warmer in the winter ... do not forget the vent if using acid cell batteries.

Hi Poor Man,
I love these kinds of projects, they're so much fun!
OK, if the light is 5 watts that means the current it draws is 5 watts divided by 120 volts or 0.042 amps. BUT LEDs don't run on AC (Alternating current like your mains power). They run on DC (direct current), so this means that inside your bulb there is a circuit to convert AC to DC to run the bulb, also the DC voltage to run the LED is much lower than the 120 volts.

That's a good thing because the LED actually uses less current than the 0.042 amps we just calculated. What makes it better is that the PV module also runs low voltage DC like the LED. So why add an inverter to boost and invert the DC from the module when you can run things without doing most of that?

The trick here for people who don't know electronics is how to disassemble the bulb and tap the output of the bulbs electronics. There are a couple of ways to do this. If you just want a light without the auto off/on then finding the two wires that run the LED is a lot easier but if you want the auto off/on the you will need a meter to test the circuit and find the point where it powers the complete circuit.

Living without, at least at the beginning of your experimenting, isn't a bad idea because you can buy boards on eBay and I guess Amazon to drive LEDs. You can still use your light fixture but what you'll do is remove the standard bulb socket and put the LED driver circuit inside with a plain old LED or a made up unit like this (this is just a suggestion, this light is 18w and is probably a spot light but you get the idea):

https://www.ebay.com/itm/2x-4INCH-36...AV9:rk:13:pf:0

That runs on 12volts DC. So you'll need a battery, a small charger to interface the battery to your PV module like this:

https://www.ebay.com/itm/12V-24V-PV-...frcectupt=true

What I would do is to get the parts I linked or something like them and a PV module.

So, how big a module do you need?
OK, if you're using a 5 watt LED for ten hours you need ~50 watt hours. (5 watts for 10 hours uses 10 times 5 watts for one hour. we use watt hours to calculate our power needs)

Let's say you have an average of 5 hours sunlight for your latitude average daily for one year. ( you can look these up for your area) If you have a 75 watt panel it will put out in 5 hours 75 x 5 or 375 wh. You want some 'overhead' because you'll also need a battery, something like an SLA (sealed lead acid). BUT the battery has something called an amp hour rating. That means how many amps it will put out for an hour but this is misleading. You really don't want to drain your battery more than about 25% to be safe and to have it last. Batteries are rated in cycles, each time you drain it and charge it up is a cylce. The deeper you drain it the fewer cycles it will have in its life. So let's say we have a cheapo 12v - 7aH, SLA battery. It has to run all night (10 hours). a 5 watt LED = 5/12 or ~416 milliamps (milliamps are amps divided by 1000; 416 mA = .416 amps, for example) just so you understand amps and watts. OK, our battery has to supply 50 wH and our battery puts out 7aH x 12v or 84wH. This is not enough because we want to only use 25% of our battery. We need a battery of 50 x 4 (which is 4 times our 25% to make 100%) or 200 wH. 200/12v = 16.6 aH thus we need a battery of this size at 12 volts for our 5 watt LED.

So there you have it. A 5 watt LED, a 75 watt module (60 will probably work), a small charger and a 16 aH 12V SLA battery.
Rob

Pool Man, there are a couple of things to consider regarding sizing of the battery and the panel.

1) How many days do you want the light to operate in the absence of sunlight?
2) How fast do you want the battery to recharge to full capacity?

I think that the first one is the most important, and that determines the battery size. Lets say that you want to turn on the lights at sunset, and leave them on for 12 hours. And you want them to be able to run for 10 days with no good sunlight.

7watts x 12 hours/day x 10 days = 840 Watt hours. Because you don't want to ruin your battery, size it for double that, and don't let it discharge below 50% capacity. So for this example, you might want two 24 series deep cycle batteries, which have about 1000 Watt hours of capacity each.

You will want a charge controller that prevents overcharging and over discharging, the two enemies that destroy batteries. The charge controllers that give the best utilization of your solar panel are called MPPT, or "Maximum Power Point Tracking."

In some cases, you can find a charge controller which also controls the lights. Morningstar makes one, although I don't think it uses MPPT for charge control, rather, the less efficient PWM.

The size of the panel depends on how fast you want to fully charge the battery after being fully discharged. Some also depends on how efficiently the charge controller uses the panel capacity to deliver the charge to the battery.

Figure that the panel can only deliver its full output for 2-3 hours per day on a good day - when the sun is shining directly at it. Before and after that, it is shining at an angle and the capacity is reduced. You will probably want it to deliver in a day, about double what the light bulb draws (84 watts hours in the above example.) If you had an efficient MPPT charge controller, you might want at least 100 Watts of panel. With a less efficient charge controller, you will need a bigger panel, maybe double that.

For what it is worth...

Bill, a 100 watt module for 2.5 hours (this is low, I believe the average for the US is 3 to 4 hours of maximum module output per day), gives 100 x 2.5 = 250 watts. A 5 watt source quadrupling the required power needed for a 10 hour night would be 200 watts, that's at ~25% battery discharge, so while, if you can afford it a bigger module is always better, you can go with a smaller module especially if space is limited.
I don't believe there's any place in the lower 48 that doesn't have sufficient sun for 10 days straight. Also 50% is too low a discharge point it reduces the max battery cycles appreciably. My home system is designed for 75 to 80%.

Two 75 aH group 24 batteries will run about $140 each, he's just lighting a 5 watt LED. His moniker is Poor Man!

Poor Man,
Here's a small cheap circuit that should turn off your LED during the day. All the parts are inexpensive and easy to get.

How it works:
The "ts3702" is a comparator, it compares the voltage on pin 2 with the voltage on pin 3. If pin 3 is higher the output pin 1 goes high and turns on the MOSFET "irfz44" which lights the LED lamp. You can change the light when the LED goes off and on by changing "R1". You can even put a variable resistor in its place but make sure to add at least a 10k resistor in series with it or you may damage the "GL5537" which is the photo resistor that senses when it's dark.
Rob

Well, a 75% cutoff would then require (4) 24 series batteries. That is too many to safely wire in parallel, in my opinion. I agree that the battery will last longer, but suggest using batteries designed for the job, not car batteries. If he can live with 5 days of autonomy, then he could use two batteries set up for a 75% cutoff. It is my understanding that these are security lights, so it is important that they stay on at night. As I'm sure that you know, a big bird splat will all but shut off the output of a small panel, so that would give him 5 days to find it and squeegee it off before his security lights no longer work.

For what it is worth, my array has not produced more than 8 or 9 Watt Hours per day for the last 11 days according to its monitoring system, and I don't expect the snow to melt off of the array for at least another week, and probably two or three weeks.

First off, he lives in Ca, a very sunny place.

Secondly, I have no idea how big or small your array is. My systems, I have two, an intertie/off line that sends excess to the grid and backs up the house in case of failures and a strictly intertie that just feeds the grid. Total array is 10.4 Kw.
I don't run off grid because it's not a good idea, I only advise running off grid when individuals don't have the option. Why run through battery charge discharge cycles when you can simply have battery back up only at times when the grid is down? I'm several hundred dollars ahead on my Electric bill and haven't paid for electricity in over ten years and that's with charging my BEV, I don't pay for gas either. I also live in a very low sun area on the 42 latitude.

But back to Poor Man's request. He's only charging one 5 watt LED, should he be spending $500 plus to do that? So, being application specific, I took into account his needs and what would be feasible financially in his location. Even if we up his battery to 12v at 30 aH he still has 360 x2.5 hours a day or 900 wH. He probably has easily twice that where he lives, so I'm being very conservative. 25% of 900 wH will give him 225 wH for his 5 watt load which is over 4 days without charging.
If we use your two 75aH batteries we have (150a x 12v) x 0.25 = 450 wH. That's 9 days to 25% discharge. The battery cost is $90 for a deep cycle AGM 30 aH battery and $215 x 2 or $430 for two 75 aH deep cycle batteries. The difference is $340 more for 4 and a half days more(I used the same brand for both calculations). It's just not worth it.

https://www.ebay.com/itm/VMAX800-Gol...frcectupt=true

https://www.ebay.com/itm/VMAX-XTR34-...frcectupt=true

"Bird splat"?

https://www.ebay.com/itm/ACOPOWER-60...frcectupt=true

Here's a 60 watt module. It's 25" x 25", that must be some big bird! The price is under $100. The total cost of the solution I gave with charger, battery, module and LED is under $200, $140 less than just your battery choice. You still need modules, a bigger charger and a lamp.

You lost me with the math. With using your recommended 25% of capacity of a 12V 30Ah battery, I get 90 Wh, only 12.8 hours for the 7 watt LED that he wants to use.

(12*30)*0.25=90

"There is an average of 3470 hours of sunlight per year (of a possible 4383) with an average of 9:30 of sunlight per day."

http://www.sacramento.climatemps.com/sunlight.php

This is for Sacramento which is more northern, not southern Ca. which has more avg sunlight.

So let's say there is substantial sun 7 hours out of that 9:30.

60w (Module size) = 420 wH per day avg. More than enough.

"The longest day of the year is 14:41 long and the shortest day is 9:18 long."

source:ibid

Using the shortest day which is dark 14.82 hours at 7w that's 103.74 wH required.

The percentage of our battery used is:

103.74/(12*30) = 28.8%

Good. We agree on one thing. The 30 Ah battery, utilizing 25% of its charge, is 90 Wh, not 225 as previously stated. 90 Wh is only enough for one single night, and will fail if there is a cloudy day.

I'm not as optimistic on the panel output. From PVWatts (a free online calculator from NREL), a December day will average 141 Wh, and a July day will average 346 Wh. Yes, it is enough to charge the battery on a perfect day for one single night of use, but just barely, and will fail to provide a single night of charge to the battery in December if there is a less-than-perfect day, or if the panel gets dirty. This assumes an MPPT charge controller, because a PWM controller would provide quite a bit less capacity.

I'm not intending for this to be a pissing match, but I'd hate for the original poster to purchase the system based on incorrect assumptions or wrong math, and then have to buy it again. It is always less expensive and more satisfying to do it right the first time.

141/60=2.35 hours of sun. He lives in Ca. I took the cap. Sacramento because it is further north. Even if there data is wrong at 9 plus hours it will never be 2.35 hours. What is the US location of your data? Where is the link? I supplied a link.

Here's another link to Ca.


"Using a yearly average, there are 5.38 daily peak sun hours across the state of California. This number is calculated for a fixed solar panel."

"California hosts a very favorable environment for solar panels, with high average peak sun hours and a lot of sunny days.'

https://www.turbinegenerator.org/solar/california/

Sacramento:
188 sunny days
partial sunny days 77
total number of days with sun 265

https://www.currentresults.com/Weath...f-sunshine.php

In the winter under worst case conditions, living in northern Ca. he still has plenty of sun.

Your advice is to use two 75 aH batteries:

12*75*2 = 1.8KwH total. At 25% = 450wH. A 7 watt ( I originally used a 5w LED because I eliminated the losses from converting from 120 to DC by using an LED that runs on 12V) will mean that he has 450/7 = 64/14.82 =4.31 days with absolutely zero sun to charge the batteries during the day. Which clearly will never be the case as 265/365= 72.6% of the time there is sun in Sacramento.

So let's look at the cost of your lighting an LED for 4.31 days of zero sunny days.

two 75aH group 24 batteries = $214 *2 = $428

https://www.ebay.com/itm/VMAX-XTR34-...frcectupt=true

To charge 1.8kw*.25 = 450w; 450/5.38 = 83.6w so we need a module that will do that:

A 100w module with a charge controller: $115

https://www.ebay.com/itm/PV-SOLAR-KI...frcectupt=true

115 + 428= $543

As opposed to

12v 30aH battery $89


100w module with charger $115

total = $204

I used same module and battery manufacturer in both assessments.

https://www.ebay.com/itm/VMAX800-Gol...frcectupt=true

Difference is $339 more It's not realistic, he's only lighting a small LED. Will there be the rare time when the system might go two days or about 50% discharge? Maybe.

And what's your solution for the LED? I included an inexpensive circuit to run a 12VDC LED. Are you using an inverter with 120v mains? If you are then you have to add the cost of an inverter and the losses from the double conversion the the $543 it already costs.

Even if I doubled the batteries it only adds another $89 to the total cost and he could do that at a later date so he doesn't have to spend for what he might not need. My design can grow, if yours is overkill what's your recourse? You already spent more than you had to and again we're only lighting a small LED.

Well, I said it before, but will say it again. The number came from PVWatts. This is the gold standard tool, from the National Renewable Energy Lab, NREL. Since you used Sacramento, so did I. I don't think it is good policy to give a wild guess when someone asks for advice.

100 watts is fine for panel size if using an MPPT charge controller. Using a cheapo is fine, too, but he'll need at least a 130 watt panel for the same charging capacity. If he wants to charge his battery faster, simply pick a larger panel.

The charge controller should also control the lights - on at dusk, off at dawn or some user defined number of run hours. It should have a proper low depth of discharge protection.

Why mess with conjuring some electronic circuit and hacking a light bulb apart, when you can just use a 12 volt bulb?

I didn't, I linked a 12V example. He doesn't take anything apart. The original bulb runs to an E26 socket which is 120V, so you need to include an inverter to 120v for your total cost.

NREL shows 5.8 to 6.5 for Sacramento. What did I quote from an alternate site:

"Using a yearly average, there are 5.38 daily peak sun hours across the state of California. This number is calculated for a fixed solar panel."

We can use MPPT but it adds equally to both analysis calculations, I didn't look back but I don't think sun hours is that critical that he needs MPPT but if he does he can substitute the module package with the CC for separates and include a cheap MPPT.

https://www.ebay.com/itm/100A-MPPT-S...DiFg:rk:7:pf:0

Please cite the cost of an MPPT CC that also controls external devices. I have expensive controllers here but they run into the hundreds of dollars.

Also starting small allows him to build to his needs. Over building leaves no reduced cost options. Look at the price your system will cost and now you want to add an MPPT with peripheral control?

He's not running an iron lung to save lives he's lighting porch light! You're up to several hundred bucks.