Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Solar panel project

  1. Jun 28, 2016 #1
    Have built a small solar panel system to power some garden LED lights. However, I am having trouble with charging the battery. Lights work for 1 night then off for 2-3 days.
    System is: 20w solar panel, 12v 30a pwm solar charge controller, 12v 18ah battery powering 6x3w LED Cree lights with 700mA flex block buck-boost DC driver. Is the solar panel too small to charge the battery in one day? Panel is in a location where it gets sun most of the day. Panel specs say it produces 1.23A but I've never been able to measure. Voltage shows 19v in full sun.
    Any help greatly appreciated.
  2. jcsd
  3. Jun 29, 2016 #2
    Oops, forgot that the lights are set to stay on for 3hrs.
  4. Jun 29, 2016 #3


    User Avatar
    Gold Member

    Let's see:
    18 watt lights, used 3 hours per day = 54 watt hours/day
    216 watt hour battery
    20 watt panel

    So if the battery were fully discharged, you could charge it in 10.8 hours, if the panel is producing 20 watts.
    And a fully charged battery would power the lights for 4 days, at 3 hours per day.

    It should work.

    Without further details, we would just be guessing what's wrong.
    How old is the battery? How many times have you fully discharged it? Do you have any 10 ohm resistors? How old is the solar panel?
  5. Jun 29, 2016 #4


    User Avatar

    Staff: Mentor

    On a perfect summer day, a 20 watt solar panel will produce about 120 watt-hours. On an average day, it should produce the 54 needed to charge the battery.

    However, I suspect that the vast mismatch between charge controller capacity and panel output is a big source of inefficiency. If your total electrical system efficiency is only 50%, then you'd need an average of 2 days of solar power to get one night of light.
  6. Jun 29, 2016 #5
    Battery is brand new. When first hooked up, it powered the lights every day for 4 or 5 days. Thought all was good so never bothered to check if the battery was discharged in the morning. Assume now it wasn't. Then it took several days to charge up the battery. Solar panel is new although I bought it off eBay so not sure of quality.

    No resistors, just using multimeter to measure voltage. Haven't been able to measure amps with it using a couple of hookups scenarios I found online.

    Also read that the amp rating of the charge controller didn't matter, was good to have more just in case you happened to need more. Not true?
    Thanks for the replies
  7. Jun 29, 2016 #6


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    It could be worth while mentioning that the panel needs to be in full Sun, without obstructions and pointing in the appropriate direction for it to perform to the seller's spec. I set up a similar system but without the 'timed' load (I use a motion sensor to turn my lights on) so my total charge draw over 24 hours is probably less. I did some pessimistic sums and came up with a 40W panel and a 14Ah battery. That seems to work in practice. I have a 'tong' Hall Effect Ammeter and have never measured the claimed maximum current output of the panel - even with full Sun (UK location). The charge controller (cheap far eastern version) seems to imply 'some' degree of charging even with low illumination but a few mA really can't make much difference to the overall charge equation.
    The lesson seems to be that my pessimistic approach to the system design was the way to go. If you want to find out how well your system is really doing then you should think in terms of actual measurement (unloaded Volts are not the whole story) - it's very easy to do an Ah calculation in your head when you read 0.25A (or whatever) on your meter and to decide whether it's supply or storage capacity that you need to increase. Those sealed Lead Acid batteries are very good value and panel prices are coming down on an almost monthly basis.
  8. Jun 29, 2016 #7


    Staff: Mentor

    That's a silly thing to say without knowledge of the OPs latitude, and the particulars of the installation. Latitude is a dominating factor.

    For reliable lighting, the panels/batteries must be sized for the worst case, not the average day. My own system uses 205 watts of panels and 440 AH of batteries to reliably provide 40 amp-hours per day of load.

    A PWM charge controller does absolutely nothing when the voltage is below 13.8. The controller is not a big source of inefficiency, it doesn't contribute inefficiency at all as long as the battery is undercharged.

    Is your battery the deep discharge type? I assume it is lead-acid.

    I find that long-lived systems with daily cycles should not plan to use more than 30% of the claimed amp-hour capacity of the battery.

    If your system has been running a long time undercharging a lead-acid battery, the battery has probably become sulphated. You can rescue it with a so-called "equalizing" charger. Otherwise you need to buy a new battery, and keep it fully charged every day to avoid sulphation.

    Although your panel gets sun all day long, it wont make much power when the sun is low in the sky. Rated watts times 5 hours per day at latitude 25 degrees is a fair assumption.

    I have no idea how efficient that 700mA flex block buck-boost DC driver is.

    Cheap self-contained solar garden lights often have just one AA battery, meaning that they run at 1.5 volts. I suspect that they run the LEDs at less than rated voltage. That gives less than maximum light output, but it saves a lot of energy. Can you adjust the output voltage of the 700mA flex block buck-boost DC driver?
  9. Jun 29, 2016 #8


    User Avatar

    Staff: Mentor

    I don't see how that can be true. The best day for any solar panel is the day that the sun is, at noon, directly perpendicular to the panel (or at its highest)....+- a few days if it isn't clear outside that day.

    Yes, I'm assuming the solar panel is reasonably well aimed.
    Agreed. My point was to explain why it might sometimes work and sometimes not: it will vary with the weather and season.
    I don't see how a charge controller could not use any power itself. But I readily admit to not having any idea how much loss there is in the charging process.

    But yes, I do get that when the battery is undercharged (as it is in this situation) the charging efficiency would be at its best.
  10. Jun 29, 2016 #9
    After 3 days of "charging" green light on charge controller was on when I got home. Does green light mean battery is fully charged or just has enough charge to be over the minimum allowed? Have had nights where the green light was on but lights only stayed on for 1/2 hour. The fact that it worked everyday for the first 4-5 days when the new battery was installed leads me to think it is not being fully charged.

    Not sure if it's deep discharge. Sealed lead acid. New battery. image.jpeg

    Panel is leaning against fence. Not rotating with the sun but never in shade most of the day. I live in British Columbia, Canada. This time of year we rarely have completely cloudy days so charging should be at its best. Winter months are unfortunately cloudy every day. Do you think I need a different panel? Would like to have this work everyday in the winter months as well.

    Anything in my system you would swap out?
  11. Jun 29, 2016 #10
    Lights just went out. On for 1 hour 35 minutes. Aargh.
  12. Jun 30, 2016 #11


    User Avatar
    Science Advisor

  13. Jun 30, 2016 #12


    Staff: Mentor

    You didn't have your thinking cap on Russ. It has nothing to do with the aiming of the panel. But this relationship between power and latitude is widely misunderstood, so I'll elaborate.

    Here are two graphics from http://earthobservatory.nasa.gov/Features/EnergyBalance/page2.php

    So we see that independent of panel aiming, the watts/m^2 of sunlight at the Earth's surface varies enormously with latitude, approaching zero at the poles.

    You may not have traveled to equatorial latitudes, and thus don't fully understand the saying "only mad dogs and Englishmen go out in the midday sun." The midday sun at the equator is much more intense than in Pennsylvania.

    You may also not have traveled to high latitudes where seasonal variations become predominant, and where the hours of daily sunlight vary as much as 0 to 24 hours per day.

    I'll end with a fun puzzle that does involve aiming the panel, but that sounds like a riddle. "For a solar panel located at the North Pole on midsummer day, how should the panel be aimed for optimum energy production?"
  14. Jun 30, 2016 #13


    Staff: Mentor

    That is a motorcycle battery. It is not the deep discharge type, and it is inappropriate for your application. It is probably heavily sulphated by now, and beyond rescue. Throw it away.

    Perhaps NiCad batteries might be best for you, but you already own a charge controller, so it might be best to buy a new deep discharge lead-acid battery. Those are the types sold for marine and RV use.

    How big should the battery and panel be? You have not told us your requirement. What is the worst case weather where you require the garden lights to work? If you live in a cold place, it might be a 4 foot snow storm and you don't shovel off your panel for a week. If you live in a warm place, it might be the number of consecutive days when you have thick stormy clouds and thus little or no solar power. Or perhaps you are happy when your garden lights work only after a sunny day.

    Finally, what is your latitude?

    You might also check out your local garden store. They might have a panel/controller/battery/light packaged solution appropriately sized for your location.
  15. Jun 30, 2016 #14


    User Avatar
    Gold Member

    We still haven't figured out why, even though all the numbers match up, the system doesn't work.
    The battery may still be viable.
    I'm currently doing an experiment, with two x 2 watt thin film panels(my preliminary analysis, is that they are garbage. The screws holding the plastic together are worth more than these pieces of........), and one of my dad's 25 year old 50 watt polycrystalline panels.

    Dad's panels are still kicking butt!
    I'm guessing his latitude is about 49.3°N, based on the population densities of British Columbia.
    Not far from me, btw. (45.5°N)
  16. Jun 30, 2016 #15


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    This conversation is interesting as far as it goes but he really needs to be measuring things. I got my Hall Effect DVM for only about 30GBP and it is useful for so many other things. The non- invasive current measuring capability is a godsend. All you need is a length of single conductor and mine will measure tens of mA repeatably
  17. Jun 30, 2016 #16


    Staff: Mentor

    Why do you say no? The picture says CYCLE on the side. It is a starting battery.
  18. Jun 30, 2016 #17


    User Avatar
    Gold Member

    I was going to say that.....
    Not sure if you have a "Harbor Freight" type company in England.
    Ours gives away meters for free.
    My younger brother goes there, just to harvest the batteries from them.
    I think I was upset that you said; "Throw it away".
    But I am with you 100%.
    I've only ever had two of these little sealed Pb batteries.
    One lasted quite a while, as it was designed for the system I used it in.
    The other, lasted a week.

    ps. Science!

  19. Jun 30, 2016 #18
    I would like them to work for 3hrs per day every day. Worst case weather is November to March where it is cloudy and rainy most days. Rarely below freezing or snow. Just cloud with little direct sun.

    Latitude is 48.4.

    Is that the same as a digital multimeter? Have that but can't get any amp reading.
  20. Jul 1, 2016 #19
    Ok, so I have the wrong battery, easy fix....does that mean that's the problem or do we still have a charging issue that needs solved no matter the battery?
  21. Jul 1, 2016 #20


    User Avatar
    Gold Member

    Actually, there is a way around that.
    Simply hook your panel directly to the LED system, without the battery.
    Then let us know what the voltage is.

    As I said earlier, theoretically, the system should work.

    That's what we're trying to figure out.
    edit: One thing I noted from the image of your battery, is that it has a rated charging capacity of 0.3 amps.
    Is your charging system limiting the amperage to that rate?
    If it is, then you're only getting about 24 watt hours of charging a day, which is half of your requirement.

    If your charging system is ignoring the 0.3 amp limit, then you may have damaged your battery. per post #22.
    I'm not really familiar with sealed lead-acid batteries.

    ps. I would share the results of my experiments from this morning, but they make absolutely no sense.
    hmmm.... Ok, I'll share them, just for entertainment value.

    Code (Text):
    initial conditions:
    __________________no load voltage_____rated watts
    black panel____________11.72____________1.5
    red panel______________22.7_____________2.0
    both panels in series__34.1_____________3.5
    dad’s panel____________17.5____________50.0
    battery(1265 wh)_______12.29(<-- indicates 50% charged)

    charging with both red and black panels in series
    conclusion: these panels can charge the battery in about .... 600 hours.

    charging with dad's 35 year old 50 watt panel
    conclusion: an hours worth of charging should add about 20 watt-hours.
    So, what doesn't make sense?

    7 hours later, after only charging the battery for an hour, the battery's voltage reads 12.47, which indicates that the battery is 75% charged.
    (75%-50%)*1265 wh = 316 watt hours.

    20 watt hours ≠ 316 watt hours.
    I'm guessing that my battery, sitting idle for the last 6 months, suffered from stratification.
    I doubt you have this problem.
    But come November, you will.
    Last edited: Jul 1, 2016
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted

Similar Threads - Solar panel project Date
Transparent solar panels? Oct 30, 2017
What's that diode in a solar panel for? Aug 28, 2017
Solar panel wires shorted togather Aug 25, 2017
Solar Panel Project: Will it Work? Jul 9, 2017
Solar Panel Charger Project Feb 5, 2013