LED flood light - Battery powered, light sensing 1

[BumpingSpheda]

Hi, I'm looking to build a Pfr to Pr converter for indoor photoperiodic plants, but don't know much about electronic circuits. I plan to power a few high power Infrared LED's from rechargeable batteries. The LED's would turn on as soon as all the other lights in the room turn off for the dark cycle (12 hours), and they would slowly dim as the batteries ran out of juice (~10 minutes). When the rest of my lights turn back on for the day cycle (12 hours) I need the LED flood light to know this and begin charging the batteries again. Rinse and repeat.

IR LED:
1.6-2.0 VDC
1000mA MAX, 900mA recommended

The light doesn't have to be "bright," in any sense of the term. I plan to run two of these ~2W LED's in series.

I realize these types of lights are made, but they require a separate, synchronized timer so they can plug into an outlet, require an LED driver, and most of them are sold for exorbitant prices. The idea doesn't seem very difficult to execute if I knew a darn thing about circuit design and electronic components (I'm thinking ~$30 max).

Any and all help would be greatly appreciated.

Best Regards,

Matt

 

[mcgyvr]

Why involve batteries/charging?
Why not just put them on a timer?

Is the slow dimming during the dark period required (wanted) or just an unwanted artifact of your use of batteries?

 

[BumpingSpheda]

Dimming is not required, but it does actually mimic more closely the spectral pattern of the sun when it sets (opposite of when it rises). Past noon, as the sun approaches the horizon, our atmosphere filters more and more Blue light away due to refraction which causes the visible light to appear more and more Orange/Red (why the sunset looks the way that it does). The IR is signaling to the plant that it needs to "go to sleep" by activating a Pfr to Pr conversion at the pigment level in the leaves. All I need to do is convert the Pfr to Pr which takes place when light saturation reaches a ratio of Infrared to Deep Red ratio that's approximately greater than 1:1. Because all the other lights are off (the plant "sees" no Deep Red) the intensity of the IR doesn't matter so long as enough Pfr is converted into Pr in a reasonable amount of time (5-10minutes) to stimulate the desired response (which normally takes place in a matter of hours if there's no IR supplementation - usually the case with indoor plants).

The reason I figured it'd be easier to use batteries is because I can move this light source to any grow room and not have to synchronize the timer, I just plug the thing in. Similarly, if I need to change the photoperiod of the plants for any reason the IR LED's will intelligently react without any input on my end.

Is the circuit design that much more complicated when using batteries?

Thanks so much for your time, btw.

 

[MacGyverS2000]

You're not likely to see the dimming nature you hope to using just batteries... LEDs have a pretty sharp turn-off voltage, so once the batteries drain out you'll see them very quickly dim and turn off (within a minute or so).

I'd stick with the timer and a wall plug. If you absolutely must have dimming, consider a slow rotation motor (like for a clock) and a circular shutter... as the disk rotates, it slowly blocks out more and more of the LEDs light.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[BumpingSpheda]

The two LED's only need ~4V, though. A 12V rechargeable with voltage regulator would still shut off very quickly? I tried to design the circuit myself, but I got stuck at the charging circuit/optic sensor portion.

If you think it's not worth the hassle, though, I guess I'll just have to synchronize the timers.

Thanks again for the input.

 

[LiteYear]

If you need mains power to run the charger anyway, the charger/battery seems unnecessarily complicated and inefficient. You can still make the switch light sensitive, but I can't see any benefit for including the battery/charger.

Use a DC power supply to power a series combination of the 2 LEDs and an appropriately sized resistor to control the current. Then, also in series, place a switch of some sort - a simple PNP transistor will do. The gate of that transistor would then be biased between a Light Dependant Resistor (LDR) and a variable resistor. This will form a circuit that switches off when light strikes the sensor and on when it doesn't. It will be finicky, very inefficient and difficult to tune. To improve on that, you might need to study a bit more electronics (or buy something ready made)!

 

[OperaHouse]

I figure if you are any kind of engineer that you have the capacity to figure things out. This sounds like a nice project foe a UNO controller (last one I bought only cost $11 shipped). Put a light sensor into one of the A/D inputs to figure when the lights go out. Then a little software program to count down time and feed that number into the PWM output. That output pin drives a FET connected to the LED. A wall wart powers the entire thing, no batteries. People in all fields have put the UNO to good use. Easy to use FREE compiler on the net so you can try before you buy. Just follow easy examples. You will likely have less trouble getting a program to work than hardwired circuits.

 

[LiteYear]

Good point Operahouse - a Uno (Arduino), Beaglebone or even a Pi is going to give you much more flexibility to tune the behaviour, and is less likely to put you in a design dead end. And I dare say there are more accessible/current/active web forums for doing this kind of thing with one of those hobbyist programmable devices, than in discrete analog electronics. Will take a bit more set up to get the programming environment going, but the rewards for persistence are great.

 

[BumpingSpheda]

Sounds good, guys. Thanks for all the advice. I'll look into Arduino. Probably a good idea for me to become familiar with it anyway as I'd like to look into 3D printing for a completely unrelated hobby.

Thanks again.