Led IS the way forwards :thumbsup:! But with out a doubt is there not enough penetration:mad: , undergrowth just doesn't get enough , you would have to have LEDs for all angles to get some REAL results. $$$$$$$$$$ :wtf:
Printable View
Led IS the way forwards :thumbsup:! But with out a doubt is there not enough penetration:mad: , undergrowth just doesn't get enough , you would have to have LEDs for all angles to get some REAL results. $$$$$$$$$$ :wtf:
To Psyz;
Having made a lighting system of roughly the power you're considering, I thought you might like to know what I found to be the most difficult part of the design, which is cooling.
A normal metal vapour lamp may produce more heat than an LED with the same number of optical watts in the PAR areas, but a metal vapour lamp can also withstand a lot more heat. A glass bulb can hit a couple of hundred degrees C without a problem, but an LED will start to lose brightness, efficiency and have a decreased lifetime if run at 60 or more.
I believe this is probably the part most people get wrong.
The biggest module I used was 50W, as a 7x7 grid of chips covered in silicone and mounted on a metal plate. I found two ways to keep the die temperature low enough. One is a cpu heatsink and fan (made of 5mm copper plate and 30 densely packed 1mm thick copper fins) which was designed to cool a 120w cpu. I ran the 12v fan fan at 7.5v as the heatsink was a little too big, so the noise level was better than full speed. The other was a natural convection heatsink (no fan) made of extruded aluminium with widely spaced fins. It measured about 20x15x5 cm. In both cases the led modules were stuck on with thermal epoxy. Small bolts, spring clips and thermal silicone grease would also work if you want them removable.
If you are a decent plumber, and have the tools and inclination, soldering (very) flat copper plates onto square tube and running a convection loop full of water and anti corrosion fluid to an external (car?) radiator would also work well if people's water cooled pc projects are anything to go by.
The bottom line is that LEDs need as much or more cooling per watt as CPUs do, because they need to run cooler to get full benefit from them.
As far as the cost, electricity usage and so on, it depends partly on where you live. All of the waste energy from the lights is heat, so if you're in a cold climate the lights' waste can be taken off the cost of your heating bill (notwithstanding the difference in price between electricity and gas), as long as the lights' coolant air (or liquid) is separate from the enclosure where the smells are.
The circuit is not one I'd recommend. It won't be very efficient, as evidenced by the large (in relation to lamp wattage) resistor. That component is to turn "spare" electricity into heat. It would be better to use a switching regulator, although I haven't looked at what dc-dc circuits and parts are available. Even the crude hack of plugging in an inverter and running mains voltage LED power packs would probably be more efficient than a linear (big resistor) regulator in many situations.
Chances are there's an ideal solution where you get LEDs in modules at voltages of multiples of minimum battery current and then use a constant current dc-dc converter with variable duty cycle pulse width modulation and solid state components.
Regarding the red/blue ratio, the figures I calculated by comparing the absorption spectra of the known light-sensitive compounds to the surface sunlight spectrum are as follows;
1 part uv
5 parts blue (439 469)
30 parts red (642)
4 parts deep red (667)
1 part infra red (735)
The wavelengths are usually rounded to the nearest 5, so the options I saw at the time were; 365nm uv, 455nm blue, 640nm red 660nm deep red 735nm infra red.
There is a problem though, because these figures are for optical watts, not electrical watts. For a complete calculation you also need the % efficiency of that wavelength and brand of LED (yep, they're all different) to work out how many electrical watts you need for each optical watt. If you're ordering cheap ones from china it's partly guesswork.
As far as the colour balance at different points in the light cycle, it may make a difference, but I've seen no evidence of that (I invite references to any material which shows otherwise). I'd suggest the best way to achieve a change in ratio is to wired up the difference between the 'seasons' as separate circuits and switch one set off and on to vary. Making the whole lot dimmable seems like more complexity and inefficiency than it's worth.
One more thing, as far as I could tell, most researchers don't believe we know every frequency plants use. The methods used to discover what frequencies are absorbed by what substance, and whether the substance is altered by that absorption, are not precise enough to detect the light spectrum equivalent of 'micronutrients'. It is entirely possible that some obscure frequency is liked by plants as a fraction too small for us to yet detect. Therefore I have assumed in previous experiments it is sensible to mix in some broad spectrum whites, such as the daylight coloured ones with yellow phosphor on a blue chip.
"The circuit is not one I'd recommend. It won't be very efficient, as evidenced by the large (in relation to lamp wattage) resistor. That component is to turn "spare" electricity into heat."
Don't be misled by the resistor wattage.
That is a "sense" resistor, and it's less than half an ohm.
That component is used to create a voltage drop.
V. = I. x R.
0.47 Ω at 2A. = 0.94V.
It is a tad oversized to limit it's heating.
Heat increases resistance.
That voltage drop is applied to the sense lead of a voltage regulator.
But, because the sense voltage tracks the current, we have tricked a cheap voltage regulator into performing current regulation.
In short, very little power is wasted in that resistor.
Less than 2 Watts at 2 Amps
P = I. x V.
2.0A. x 0.94V. = 1.88W.
This regulator circuit works very well with battery power.
Just a couple 'lectronic pennies.
Aloha,
Weezard
2 Watts wastage seems acceptable, the more leds you had on it the less significant that would be.Quote:
Originally Posted by Weezard
Now I look at the article properly I can see there are various different circuits described, the first being more in the way of an introductory "what not to do". Mainly I was just prejudiced by the unusually large resistor, should have read the whole thing first.
The second to last configuration reminds me of something I used to drive the backlight leds on a phone screen.
Very interesting, despite unfair first impression.
hi so what kind of light do you use currently?Quote:
Originally Posted by khyberkitsune
i have dumb quetion. . .(idk anything about lamps)
what is nm ???is that size of the LED lamp?Quote:
940nm, 850nm, 660nm, 630nm, 610nm, 590nm
570nm, 530nm, 510nm, 470nm, 460nm, 400nm
whats differences between normal led and full spectrum led?Quote:
for good growing you need blue, red and deep red spectrum.
can i just use normal LED Lights?
So the most recent post dates back to 8/2012 - Where is current LED technology as it relates to MJ growth? Any recommendations?
PPP-Dragon
Um, partial spectrum/full spectrum and you are correct, these questions do border on dumb.Quote:
Originally Posted by MG.Snake
Please read.
Your last question is well answered right here, all you need do, is read it, yah?
Aloha and good luck,
Weezard