joesmithjoe
Member
Let's take a look at this for a second, and examine what a typical panel will draw, based on the chips used.
Around ~97.3% of all LED panels (within a Standard Deviation of ±1.34%) that people are using for growing...originate in China from the same ~dozen factories, all employing Epistar, SemiLED, and Bridgelux chips, or a close variation of.
(See, folks? 'Statistics'...)
In figuring a rough, estimated draw, my first guess was 'it's probably going to be around 70% red, 20% blue-green, and 10% IR')...
...and after blowing up a close-up of the Bloom panel and counting the actual LEDs used, turns out this # was accurate. Here's what I got from the ol' eyeballs:
- 200 Red
- 58 Blue/Green (16 green, 42 blue)
- 30 Infrared
(By the way, using green in an LED panel...is a dumb idea.)
By the way, GSL's site lists the same 550w draw for both their 'full spectrum' and their 'bloom' panel - even though a typical IR chip uses less than half the volts as a blue, green, or white, at the same current.
Very improbable - unless you're also taking out additional reds and back-filling with a precise number of blue and white, for example.
---------------
Most of the typical panels used around this time where the 'designers' (i.e. Marketers, usually with little or no actual knowledge of either LEDs - or plant physiology, for that matter - who were simply submitting RFQs to China manufacturers for drop-shipping) and were pushing the thermal management capabilities of these '3w' designs to their limits (eventually curtailed by the engineers, who know better, hopefully)...generally ran them @ ~500mA, which is ~43% higher than typical '1w' levels...in the same 1w boards using similar heatsinks and fans. Bad idea.
(The Grow LED Hydro/'Spectra' 2010 panels, etc. are a good example of this).
Using these chips, here is about what you'd expect to draw at this current, including for the eight fans used in the design, which employs a standard 12x24 (i.e. 288 chip) board:
View Image
So, using the typical emitter packages employed by these factories, at 500mA, you'd be getting something pretty close to what you've found, B.
Depending on how 'dirty' the power grid is in your area, your intra-day readings may vary within several watts, but no more than that.
--------------
Ok, so that's pretty typical, with these low-to-mid-grade chips. Now, if this panel was using good chips (CREE, OSRAM, LUXEON, etc.), and drawing roughly the same power from the circuit, what current would they have to run at in order to achieve this?
View Image
Using good, 1st-grade chips, one of the things you'll notice is that not only are they able to run at a lower forward voltage (Vf) than their competitors at the same current (a characteristic of all good chips), they also have less voltage drift as current increases - which makes them markedly more efficient, too.
Note that you're able to run these chips at ~70mA more (570 vs. 500), and still pull the same number of watts, which is impressive.
(Incidentally, if you're running all the LEDs on the same driver, which most of these models do, 570mA is about the highest current you'd want to run any particular string at, due to the performance characteristics of red LEDs.)
------------
So with a better chip, with this particular example, one might also infer that it's possible to:
- Run them at the same current (500mA), and use almost 50 fewer watts (!) (and a lot less heat), and still get more usable light, vs. your typical Epistar/SemiLEDs; or
- Run them at the same power (405-410w), and get TONS more light! (and still get less heat, due to increased conversion efficiency)
-------------------
Ok, so...what if they were actually running these panels at "3w" (i.e. 700mA), "Balls To The Wall" levels - a decision which would rank right up there with going into the ghetto at night,starting a land war in Asia, and Lindsay Lohan's acting turn in Herbie Fully Loaded?
View Image
(Note the forward voltage creep on these LEDs @ 700mA, vs. CREE/LUXEON - i.e. 3.8v vs ~3.3-3.4v for white/blue/green, and 2.7v vs. ~2.3v for red. That's a HUGE difference.)
Hmmm...apart from being able to fry an egg on the heatsink and having sparks spittin' off in all directions and setting the curtains on fire (which "all good growrooms have in 'em"...right?), looks like you'd get right around 600w, while it lasted.
------------
Just be glad you're not actually drawing more - otherwise, that would be horrible for your thermal management, and thus your longevity.
But...yeah, they really need to update their site.
Here, let me help...I've got this ball peen hammer handy that should just about 'do the trick'.
Cheers,
-SX
I have two lighthouse blackstar 180w pannels with the removable cord. If you were to replace the thin aluminum panel in them with a nice thick heatsink, could you up the amperage and have them live? If for example they pulled 500ma and you upped it to 700ma would they live as long as they are kept cool enough?