I am making a little dusk-to-dawn setup for a light box with my house numbers in it to go on the mailbox post. They're the flicker/torch/candle type LEDs.
The bulbs are advertised at 6W each, (two bulbs) but that is the sum of all the chips. It likely consumes less than 6W when it's flickering, but we'll call it 12w/1A total to be safe.
I want them to be able to be lit for 12 hours, so that's 12Ah, but I'm assuming I should probably go a little bigger? I'm sure the charge controller likely kills the load at 11v or so.
Then I would technically need a 12w panel, but since I live in PA and cloudy and snowy things happen, 20w? Is that enough?
What am I missing? Seems like the battery is pretty big
If you already have the bulbs, I'd stick an ammeter in there and see what your actual draw is. I suspect it's quite a bit less than 1A and that ends up being important.
12AH batteries are basically Power Wheel batteries, you can get AGM ones for under $30. Note that if you're expecting even one day without sun, that will double your required battery capacity. If you want to handle a week, now the battery gets big!
The effectiveness of your panel is going to depend on if it gets clear sunlight (no shadows! they hate that, even partial ones) and if you have it tilted towards the sun. With snow, it'll drop pretty much to 0 output unless you brush it off when you plow the driveway, so it all depends on how tolerant you want your rig to be of snow etc.
It is, above all, a decoration. If it fails once or twice I'm not going to lose sleep over it. Any charge controller I get will have under-volt shut off, so the battery should be protected. Even if we get some cold/snowy streaks, the battery should hold enough charge to prevent freezing even if the lights don't work for a few days. I want to assemble this and two years later realize that the lights aren't working and I replace a battery or something. I just don't want the battery to be the weak point because I didn't give it enough charge.
I don't have the bulbs yet, but an ammeter would be a wise call.
I'm tempted (since a 20w panel is pretty big) to get individual 2x2 cells that I can attach to the top of the mailbox instead of having a huge panel.
I'm also re-evaluating my goals as I shop for 12Ah batteries. Maybe I choose a smaller-wattage bulb, or maybe I put up with a shorter run time each night.
In reply to Keith Tanner :
Does the rest of the math look right?
I don't have any real depth of knowledge in the math. But looking at my own array, I get the equivalent of about 8.5 hours of peak output at this time of year when it's 15 hours from sunrise to sunset. That's because the output follows a bit of a bell curve over the course of the day. Mid-winter, it's obviously a lot worse. These are panels that are basically sitting flat instead of being angled to the sun, so they could be a bit better but you're sure not going to get 100% the entire time the sun is out.
Can I make a very left field suggestion? UV reactive or glow in the dark paint?
I've also got some powdered additive that lasts for several hours. Maybe not fully overnight, but definitely until last call. When it's mixed with silicone, I haven't tried paint personally.
In reply to RevRico :
Interesting thought. I think there is enough ambient light that it won't glow well enough. It's one of the reasons I'm using 12w worth of LEDs so it's bright enough.
It's pretty dark. I'm in the middle of a suburban block with street lamps at each corner, so dark-ish.
The basic math of 1 amp for 12 hours == 12 amp-hours is fine.
Determining how much panel you need to charge a 12 amp-hour battery in a day's worth of sunshine is more complicated, which is what Keith is getting at. It depends on your latitude, the angle the panel are mounted, the efficiency of the charge controller (PWM or MPPT), how often the panels are shaded (any trees nearby?), how much bird crap or other dirt is on them, the amount of cloud cover, and a whole bunch of other things.
In reply to Keith Tanner :
Holy crap 116.5 kWh? What's your daily usage because that's about 4 times my usage in this older home.
In reply to Stampie :
That's pretty close to peak output, it maxes out at 125 or so. Six months later in January, it can only manage half that. It's actually kinda hard to figure out daily usage from logs because that's not measured directly, but we are net positive or equal on production for at about 9 months of the year.
The array ended up being more efficient than expected and I found a few ways to tweak it so it's overproducing. We are feeding the neighborhood with power :) Also means that if we get another EV, it will cost us nothing extra to fuel.
Well, I ordered stuff. We'll see how it goes.
I chose a 15w panel. It may not be enough, but we'll see. I chose it partly because it was a name I knew from watching Will Prowse videos, and also because it was the perfect size to live on top of the mailbox like a hat. I got a 15Ah battery, a Schumacher charge controller, a dusk/dawn sensor, and some cable glands for an ABS waterproof box.
Now if only my plasma torch would stop being a dick, I could finish cutting the numbers for the address.
In reply to codrus (Forum Supporter) :
One tree nearby, but the arc of the sun rarely puts shade on the mailbox. Not sure about the charge controller. It's a Schumacher SPC-7A. I can install the panel at any angle. I'm sure there is a website that says "for Pennsylvania, angle it X degrees south."
The Victron charge controllers offer the option to basically run "dusk to dawn" lights based on solar input. They can also do timed on/off through their existing output, so a total one stop solution.
But they are likely out of the McBudget.
Time of the year and direction of the panel make a huge difference in power output. The NREL PVWATTS calculator is a good resource for figuring out what your panel will put out. Actual measurements and calculations would be nice, but since you have stuff on order the right answer is probably not to overthink it and just try what you have.
In reply to jwagner (Forum Supporter) :
But I'm SOOOO good at overthinking.
Most of the parts are arriving in the next few days, so I'll be back to update.
My only concern for your system is the battery. Lead acids don't like to be drawn below 50% charge. Lithium can go 80% but don't like cold weather.
If you are pulling 12 ah off of a 15 ah battery you are exceeding the 50% rule. It may shorten your battery life.
In reply to Toyman! :
The charge controller turns off the load at 11V to prevent discharging too far. I think the issue will be how quickly it drops to 11V.
In reply to Curtis73 (Forum Supporter) :
Assuming you are using a small AGM, 11 volts may be too low depending on your load. At low/no load, an AGM is down 50% at 11.98 volts. 0% is at 10.8 volts.
If the cutoff voltage is adjustable, I'd set it to 12 volts and see how long the battery runs. Then check the no-load voltage after sitting for 30-45 minutes. If it stabilizes above 12 volts then you can turn your cutoff voltage down some and try 11.5 volts or lower.
You want to end up with the resting voltage right at 12 volts after running your load to the cut off voltage.
In reply to Toyman! :
I will admit to not being as knowledgeable as you, but I was going off this spec sheet from the battery which lists 10.5V. And yes, SLA/AGM
Correct. But, 10.5 volts is dead. While the battery can do that, it is hard on it and shortens battery life significantly.
Take my father's golf cart. He insists on running it dead before charging. He gets a year out of a set of batteries. If he would recharge nightly when it is above 50% discharge, he would get 4-5 years out of the batteries. A set of batteries is about $1000. I've explained it to him countless times and he refuses to understand.
If you recharge at 11 volts, that's probably around 75% discharged. It's going to shorten the life of a battery. It may not be an issue if the battery is fairly inexpensive.
Oh, and do a build thread, I'd love to see what you end up with. It sounds cool.
Toyman! said:
Correct. But, 10.5 volts is dead. While the battery can do that, it is hard on it and shortens battery life significantly.
Take my father's golf cart. He insists on running it dead before charging. He gets a year out of a set of batteries. If he would recharge nightly when it is above 50% discharge, he would get 4-5 years out of the batteries. A set of batteries is about $1000. I've explained it to him countless times and he refuses to understand.
If you recharge at 11 volts, that's probably around 75% discharged. It's going to shorten the life of a battery. It may not be an issue if the battery is fairly inexpensive.
Good to know. I wasn't sure how Ah was measured and at what voltage threshold.
I think at this point I'll first try hooking up one bulb to see if it makes enough light. That would at least halve the load. If it's not enough light, I'll just do some testing the first few days to verify the charge/discharge cycle. I think I have an old VDO voltmeter that one of you sent me for the New Year's game and I might just hook it up for a while to make it easy to check.
Not enough for a whole build thread, and I talked a little bit about it in the GRMFB thread, but I started with an existing 36"L by 4x4" by 3/16" wall post that I scavenged from another theater when they remodeled and took out a safety rail on their loading dock. I cut some "leaves" with a plasma torch and then drew vines with a grinder. I welded a 1/4" steel plate to the top to mount the mailbox.
36" isn't quite enough height for USPS. They suggest 41" - 45" above the street. My lawn goes a couple inches up, so I have a plan shown below to get four more inches of height.
The post will get sandblasted, and I'll fill the vines with paint and allow the rest to rust.
I wish I knew why it copied and pasted in black, but oh well.
The green is grass. The street is on the left. The hole will obviously be filled with concrete. The red box is a 13x13 hardieboard box with an open ends that gets filled with 3-4" of more concrete with bolts set in it. The post will bolt down to those. The hardieboard box will be covered with leftover stone veneer from my fireplace project. The box off the back of the post is my house number, and the small gray box near the ground is the box with the battery and charge controller in it. PE panel will go on top of the mailbox, and all the wiring will run inside the post.
This is as far as I got on the post. My plasma torch decided to quit working and the "protection" light is on, so I have to fix that before I can cut the light box with my house number on it.
This picture shows the post, and part of the hardieboard plinth/box.
Once it's all in place, I'll fill the void between the stone-veneer hardie board box and the post with dirt and plant something like Phlox in it. Something hardy that doesn't get big and flowers periodically. The nice thing about 3/16" steel walls is that I'm not worried about rust-through for about 30 years, but I might powdercoat the bottom 9" where it's in direct contact with dirt.
That's going to look awesome!
Getting close to done. I have to wire it and install it yet.
