Floodlight flickering

[joelonly]

Anyone can help me? i have 50W off grid solar system, 12V/30W LED floodlight directly connected to the solar controller. 50Ah deep cycle battery.
New battery only lasted for 6 months, the LED floodlight flickering. What is the possible problem
Thanks

Joelonly

[Naptown]

Bad charge controller?
Not really enough info to determine.
Could be a bad panel
Undersized panel for the load
Insufficient insolation to charge battery

[Mike90250]

My bet is on the LED driver circuit being bad, overheating, or a bad wire connection. Flicker is not caused by battery, unless it's a rare intermittent connection in the cells.

It could be the solar controller LOAD terminal is overloaded by a 30W load too.

[PNjunction]

This is interesting. Made me pull out the pencils for napkin-calc and a propose better solution than the dc bulb, and use an AC inverter with an ac led flood instead! Bear with me as I geek out..

I looked up a 30w DC driven led flood (golden gadgets) and see that true enough, it pulls 30w. Rated at 2.75a current at 12v dc.

However, if you were to run an "equivalent" 30w ac flood led, (GE PAR38 typical) that only pulls 12w in reality, and using a high-efficiency low wattage inverter, even a simple Samlex MSW, that would be about 12w/10 (samlex formula) = 1.2a at the terminals. That's a bit more than 50% better than the dc-driven led, even with the inverter!

I never thought about that as I just assumed that dc-driven bulbs were more efficient than ac driven ones. Wow.

You may want to look into that. Plus we have some issues from the napkin calc:

((50ah batt * 12v) / 2) = 300 usable watts reaching 50% DOD.

If you use that 30w flood at night for 10 hours, you are done right at 10 hours.

So now you have 25ah (plus about 10-15% to account for recharge inefficiency) - lets call it 28ah needed to return daily....

50w panel / 18v = 2.7A under absolute best conditions....

This would take 28 / 2.7 = 10.3 hours of charging daily, which nobody has in real-world terms of solar insolation. (differs from just sunrise to sunset)

If this is the case of running it nightly for 8 or more hours, I'd DEFINITELY beef up your panel to at least 100 watts or look into the AC driven led bulb that is actually more efficient with a *quality* high-efficiency low wattage inverter.

[PNjunction]

Sorry about this - I think I just figured it out.

You are using the controller's load terminals to program the light on/off durations.

That in itself may be an issue at 2.7a draw from the load terminals. If so, then use of a relay which then controls the dc bulb and still allow for programmable lighting duration may be in order.

OR, if the load terminals can handle that, it may just be a simple matter of your battery not getting charged enough daily, and the voltage is too low for the dc bulb to operate without flickering. If you've been doing this for a long time, the battery may be sulfated, complicating matters further.

Grab a voltmeter, and measure the voltage of the battery terminals when it is flickering.

I've got a feeling that your best bet regardless is to double the size of your paneling, and of course make sure your CC can handle that.

While my previous suggestion to use an AC led bulb and a quality simple inverter would definitely save you power, you'd lose the automation and have to turn off the bulb and inverter manually. Forget to turn it off, or just turn off the bulb and not the inverter, would drag the battery down to 10.7v typically before the auto cutoff activates - which is not a good thing to do.

[joelonly]

Bad charge controller?
Not really enough info to determine.
Could be a bad panel
Undersized panel for the load
Insufficient insolation to charge battery

Thanks

[joelonly]

Sorry about this - I think I just figured it out.

You are using the controller's load terminals to program the light on/off durations.

That in itself may be an issue at 2.7a draw from the load terminals. If so, then use of a relay which then controls the dc bulb and still allow for programmable lighting duration may be in order.

OR, if the load terminals can handle that, it may just be a simple matter of your battery not getting charged enough daily, and the voltage is too low for the dc bulb to operate without flickering. If you've been doing this for a long time, the battery may be sulfated, complicating matters further.

Grab a voltmeter, and measure the voltage of the battery terminals when it is flickering.

I've got a feeling that your best bet regardless is to double the size of your paneling, and of course make sure your CC can handle that.

While my previous suggestion to use an AC led bulb and a quality simple inverter would definitely save you power, you'd lose the automation and have to turn off the bulb and inverter manually. Forget to turn it off, or just turn off the bulb and not the inverter, would drag the battery down to 10.7v typically before the auto cutoff activates - which is not a good thing to do.

Interesting, Thanks

[joelonly]

Sorry about this - I think I just figured it out.

You are using the controller's load terminals to program the light on/off durations.

That in itself may be an issue at 2.7a draw from the load terminals. If so, then use of a relay which then controls the dc bulb and still allow for programmable lighting duration may be in order.

OR, if the load terminals can handle that, it may just be a simple matter of your battery not getting charged enough daily, and the voltage is too low for the dc bulb to operate without flickering. If you've been doing this for a long time, the battery may be sulfated, complicating matters further.

Grab a voltmeter, and measure the voltage of the battery terminals when it is flickering.

I've got a feeling that your best bet regardless is to double the size of your paneling, and of course make sure your CC can handle that.

While my previous suggestion to use an AC led bulb and a quality simple inverter would definitely save you power, you'd lose the automation and have to turn off the bulb and inverter manually. Forget to turn it off, or just turn off the bulb and not the inverter, would drag the battery down to 10.7v typically before the auto cutoff activates - which is not a good thing to do.

This is interesting. Made me pull out the pencils for napkin-calc and a propose better solution than the dc bulb, and use an AC inverter with an ac led flood instead! Bear with me as I geek out..

I looked up a 30w DC driven led flood (golden gadgets) and see that true enough, it pulls 30w. Rated at 2.75a current at 12v dc.

However, if you were to run an "equivalent" 30w ac flood led, (GE PAR38 typical) that only pulls 12w in reality, and using a high-efficiency low wattage inverter, even a simple Samlex MSW, that would be about 12w/10 (samlex formula) = 1.2a at the terminals. That's a bit more than 50% better than the dc-driven led, even with the inverter!

I never thought about that as I just assumed that dc-driven bulbs were more efficient than ac driven ones. Wow.

You may want to look into that. Plus we have some issues from the napkin calc:

((50ah batt * 12v) / 2) = 300 usable watts reaching 50% DOD.

If you use that 30w flood at night for 10 hours, you are done right at 10 hours.

So now you have 25ah (plus about 10-15% to account for recharge inefficiency) - lets call it 28ah needed to return daily....

50w panel / 18v = 2.7A under absolute best conditions....

This would take 28 / 2.7 = 10.3 hours of charging daily, which nobody has in real-world terms of solar insolation. (differs from just sunrise to sunset)

If this is the case of running it nightly for 8 or more hours, I'd DEFINITELY beef up your panel to at least 100 watts or look into the AC driven led bulb that is actually more efficient with a *quality* high-efficiency low wattage inverter.

Thanks

[joelonly]

My bet is on the LED driver circuit being bad, overheating, or a bad wire connection. Flicker is not caused by battery, unless it's a rare intermittent connection in the cells.

It could be the solar controller LOAD terminal is overloaded by a 30W load too.

50Ah battery only last for 6 months and started to flicker. But, it is ok if directly connected to the battery

[Naptown]

Thst narrows things down a bit.
What is the model of the chsrge controller and panel where is the system and how many hours a day is the light on.

[Bucho]

I looked up a 30w DC driven led flood (golden gadgets) and see that true enough, it pulls 30w. Rated at 2.75a current at 12v dc.

However, if you were to run an "equivalent" 30w ac flood led, (GE PAR38 typical) that only pulls 12w in reality, and using a high-efficiency low wattage inverter, even a simple Samlex MSW, that would be about 12w/10 (samlex formula) = 1.2a at the terminals. That's a bit more than 50% better than the dc-driven led, even with the inverter!

I never thought about that as I just assumed that dc-driven bulbs were more efficient than ac driven ones. Wow.

I'm skeptical of this, especially since the "AC" LED's I've bought thus far are actually DC LED's. Are you sure you're comparing apples to apples?

[sensij]

I'm skeptical of this, especially since the "AC" LED's I've bought thus far are actually DC LED's. Are you sure you're comparing apples to apples?

It is definitely not apples to apples. A 12 W GE PAR38 puts out 820-1050 lumens, depending on the model. The 30 W golden gadget light puts out 2200 lumen.

The 30 W DC light claims to be 150 W halogen equivalent. The GE light claims to be a 90 or 100 W incandecent replacement. They are not the same thing, and if you look at the numbers, you can see there is no magic efficiency gain from an AC LED.

[Bucho]

It is definitely not apples to apples. A 12 W GE PAR38 puts out 820-1050 lumens, depending on the model. The 30 W golden gadget light puts out 2200 lumen.

Yeah that's what I was seeing.

[Naptown]

At what distance and beam spread.
Led's tend to be directional

[PNjunction]

Aha - thanks for the heads up on lumens - I didn't consider that, which being a flashlight guy you'd think would be at the top of my priority list.

But that does raise the question - is part of those 2200 lumens being wasted in some direction that you wouldn't want in a flood? To be fair to both, what is the testing criteria that each follow before they rate them for sale? Is it the same?

This would be a good thing to find out considering the amount of power one can save if the measuring / rating procedures are different among the manufacturers. And as we know, the marketing departments can hold sway over the engineering departments.

Time for some nighttime beam-shot and beam-spread pics.

Seriously though thanks for the heads up - they are not apples to apples if the light isn't in a pattern you don't need / like.