# Off grid LED solar lighting without batteries

I have 2 COB LEDs (12-14V DC, 50W, I think with 3,75 A in data sheet). I need to run them with a solar panel, it should only run when there is sunlight no batteries needed. Using a 270W panel (open circuit voltage around 38,8V, optimum volt and current: 32V, 8,4A) with 2 parallel connected DC DC Step Down Buck Converters 8-40V 10 or 12A Constant, adjustable current and voltage to run them both separately, it is concerning because the maximum voltage of the solar panel gets really close to the maximum current the Buck Converters can handle.

Using only one LED with a 130 W Solar panel (open circ. volt 22,8V, opt. volt+curr: 17,6V 7,4A) with a 7-32V DC DC step down buck Converter 8-10A also constant adjustable current and voltage. It is concerning that the input voltage may get to low for the converter (because just 38% of the current of the max. output is needed and because the voltages get lower when the temperature gets over the 25°C Standard-conditions.) Which means bad efficiency when there is no perfect sunlight shining.

Or are my concerns unfounded

I got the idea to try it with a buck/boost converter (link) to hopefully solve the voltage problem that comes with a 150W Mono solar cell at higher temperatures (shift down to almost 12V in the worst case (75°C and 0,3-0,4% irridiance)), what do you guys think does it work better that way than with a buck converter? So in this way: Solar panel -> Buck Boos converter -> LED

I mean the Buck boost converter only costs about 3$so... Instead of a Buck converter. higher voltage, lower amps mean less heat, so there Comes another Advantage with a buck boost converter i guess, because I can go higher with the voltage as I thought (13,5 instead of 12,5V with the buck converter to slightly better handle the temperature-caused voltage Drops) And for the Buck Boost Converter it should be of no or little concern if 13,5 or 12,5V are adjusted because it can both, also a Switch of the converter just stays opened when it gets the same vin as vout is adjusted, am i right? just the current gets witheld in that case) OK thank you i now ordered a Automatic Buck-boost converter but you are right a mppt Charge Controller (that i have to adjust?or is it enough to connect the battery first with the chargecontroller? ) and a small battery would be better for more continous LED lightning. I found this one https://www.ebay.de/itm/12V-24V-30A-MPPT-Solar-Charge-Controller-Solar-Panel-Battery-Regulator-Dual-USB-/283300643163 ... In combination with a small 12V 20AH solarbattery..? But first i will try it Out without battery so heres my First Plan: So do you think it works out when i Just wire the 150 W Solarpanel (21,8V opencirc. V) with the Buck-boost converter (How about the polarity of the voltage with buck-boost converters i think i did not fully understood this till now will/can it work too with a combinated step Up/down module instead of a normal Buck converter?) and then to the one cob led that is mounted to an Aluminium heatsink? I also ordered two Buck converters to experiment. Here are the links to all the articles that i ordered to make it complete: BUCK BOOST CONVERTER 5-30V https://www.ebay.de/itm/Automatic-Boost-Buck-Converter-CC-CV-5-30V-To-1-25-30V-8A-12V-24V-Regulator-100W-/132951586178?txnId=1566909161003 • If you only want to power 100W of LED from 270W of solar panel, efficiency isn't really a concern. Even at half its rated output you still only need about 75% and most converters can manage that easily. – Finbarr Mar 11 at 11:15 • Note that while you don't necessarily need batteries in your application, your cost effect would likely be vastly improved with a Solar-->MPPT charger-->Battery-->Constant current LED driver(boost/buck or similar topology if necessary)-->LED arrangement. MPPT optimizes panel output at a given light level, resulting in intermittent charge voltage to the battery(or capacitor bank). The battery(or cap) need only be large enough to average the output. – K H Mar 14 at 0:55 • If you were feeling super ambitious and willing to learn a lot, you could design something like an MPPT Cuk converter with its output voltage varying instead of it's duty cycle(off the shelf products provide constant output voltage at varying duty cycle for battery charging). This would allow you to achieve higher efficiency without the addition of the battery. The nature of the cuk converter should allow the LED to DIM all the way to 0% as the input power drops, providing the most useful possible range of light. This would also require output current limiting for the LED. – K H Mar 14 at 1:02 • You want to turn light, into electricity, so you can turn it into light. Why not just use a Solatube or similar? – TimWescott Aug 13 at 22:04 ## 3 Answers Or are my concerns unfounded Well, of course you have to consider the maximum input voltage your circuit can handle. Regarding your concern that the input voltage may get to low for the Converter we can say: Sure. But this is actually only a question of system dimensioning: A certain panel can supply a certain peak power at a certain sunlight input. If you want to drive your 2x50W LEDs at full power, you have to connect as much panels as necessary to get that >100W at a certain sunlight intensity. If you want full LED power at dusk and dawn, you will need more panels - but you can't use their potential power at noon if your LEDs are the only consumer. If you want full power only at noon, one 130W panel for both LEDs may be enough. Boost or buck converters may not be a good choice for direct-from-sun applications. They draw more current as the input voltage gets lower. Hence, at the instant when the solar panel does (even for a brief moment) not provide enough power to drive your LEDs at the (maximum!) power you set the converters will quickly spiral the voltage down and your LEDs turn off. Yes, even if the panel only sags to 98% of the LEDs' power demand, the LEDs will turn off. Also be prepared that a solar panel is much more sensitive to changes in insolation than the human eye. You may find that you hardly notice the difference between 100% and 50% absolute insolation while for the solar panel it already makes 50% of a difference. Don't be disappointed if you only get 10-50% of the rated power out of a solar panel on a seemingly nice sunny day; consider that when dimensioning your system. Maybe do some tests (with smaller panels) and see how much of the rated power you get in weather conditions you consider somewhat average for your location/time of year/day. Answer to 1 concern. More later maybe. If Vpanel gets too high you can add a load - in several stages if desired. A zener driving a Schmitt trigger (6 in a package for$1) with a light bulb load.

Slightly more sophisticated. Slightly.
Feed panel via a diode to a modest capacitor. If Vcap reaches a limit short panel (on panel side of diode) with a MOSFET until cap reaches some slightly lower voltage. Then remove short and repeat. In full sun at low load the FET will cycle on and off. A panel with say 10A Isc and a 10 milliOhm Rdson will dissipate power = I^@R = 100 x 0.01 = 1 Watt in the FET if always on and less to much less usually.

A buck converter can be arranged to NOT aim at a target Vout but at a target Vpanel ~= Vmp. This is a rough MPPT converter. If Pload > Ppanel it does the best it can. If Pload << Ppanel Vcap rises and the input shorting FET takes over. You now have a system that MAY produce too much V on light loads but a little more playing with inverter control will give you what you want,

If Vcap_panel > upper_limit turn on shorting FET If Vcap_panel > lower_limit turn off shorting FET

Combine: Maintain Vp at Vmp and If Vout > Vdesired stop converter.
If Vout < Vdesired - tough luck.

______________________________

Odds are the Maintain Vp at Vmp is not needed, ie If Vcap_panel > upper_limit turn on shorting FET If Vcap_panel > lower_limit turn off shorting FET If Vout > Vdesired stop converter.
If Vout < Vdesired - tough luck.