Apollo Series LED panels from Cidly

thats how i felt.. its like so you dont want pictures of your product out there.. wtf.. if i was to buy an apollo would he ask me to take down pics if i did after i bought it? kinda shady..

shady indeed whats he got to hide?

Yeah wtf does he have to hide....I mean they all copy each other over there anyways...

You can't patent particular wavelengths in a board, folks. That's just marketing BS. No attorney worth his salt would touch it with a 3m pole.

I welcome the first idiot drop-shipper that tries to make such a 'patent infringement' stick - on a Shenzhen OEM, or the buyer.

Originally Posted by ChrisGGG

@ The Lurker

I just meant clone the EVO in the sense of using just 10000k and 660nm not the higher quality leds. That link to e.shine was pretty cool. How can we be sure they are using real Crees? I would use a less efficient led if it cost 1/5 and was 2/3 the efficiency, but I guess that's the problem with knock-offs you can't trust the spec sheets.

Gotcha; no worries. One must consider lumens/dollar as well as lumens/watt. It might not be as efficient, but neither does it mean that it has to be crap, too.

In e.shine's case, after further checking in the reef/coral community, it turns out they have actual CREEs available, as well as Bridgelux for ~35-40% less. Their Aqua Washers look promising, too.

But if a drop-shipper (HGL, Advanced, Pro Source, etc.) tells you they're using CREEs in their panels, take that with a grain of salt.

I may decide to pull the trigger on some of those myself later this year, once my situation is more settled.

Originally Posted by Psytranceorgy

yeah I wish Cidly would still offer 60 myself, well I guess maybe they do, but you have to buy a Blackstar Chrome from Lighthouse/Gotham? I dunno... but I am going to go all 90 and I've come to terms with it...

Don't worry about it Psy; just use the 90's and grow shorter plants. With 60's you'll always have something still in shadow, unless it's one-plant-per. Better to have more consistent coverage and a larger light footprint; 60's are overrated.

Originally Posted by Psytranceorgy

So it sounds to me like Sing is able to simply reduce the input voltage (via priprietary method) to the individual RED/ORANGE LEDs from 2.7 to 2.6... independent of input current...

Cheers!

Thanks mate! Yes, there are several ways to reduce voltage to a string or partial (sounds like he'll be running a separate one off the same driver for just the reds - always better to have one driver per string instead), it's Electronics 101; problem is such methods always reduce efficiency to a degree. But as long as it works for ya, then hey, you'll get a working light out of it.

Originally Posted by squarefodder

Iwill ask for the exact specs.

I did go back and see where Psy posted the cool white @ 84.9 lumens per watt. So an apollo unit drawing around 200w would mean 16980 lumens correct? Maybe not because not all wattage is going to the led.

Unfortunately sf.....no, not in actual practice. If we wanted to be more precise, then we'd have to take into account both 1) Driver Losses (~10-20%), AND 2) Temperature De-rating at the junction (another ~10-15%):

CREE_XP-E_White_Temp_Derating.png

(This graph is from the CREE XP-E specifications. I expect an Epistar to be less efficient still.)

Note that at 700mA+, it's common to have a Tj between 70-100°C--! Don't be fooled by the temp on the heatsink exterior or the fixture; junction temperatures (which dictate emitter efficiency) are always much higher.

(Right now, almost all spec sheets are rated at a consistent Tj of 25°C. You will NEVER get this low in-circuit; that's lower than most ambient air temps just in your garden! Only way you'd get even close to that is using Peltier cooling (which is massively inefficient, and uses more than twice the watts it employs for cooling) or with a water-cooled HS, like you see in some high-end gaming desktops.

Most folks forget about de-rating performance when estimating output, simply because they don't fully understand how LEDs and typical electronic/solid state circuit losses work. But unless they've got a background in it, they shouldn't be expected to.)

--------------

With typical drivers used in these units, expect no better than ~85-86% efficiency at operating load.

So, if you were using their (Epi) cool whites (not warm), and those spec sheets were actually close to real-world performance, you might expect to actually get something closer to:

200 * 0.85 * 0.85 x 84.9 = 12,268 lumens, or thereabouts.


To get ~200w at the LEDs, you'd need an initial draw from the wall of closer to ~235w.

(With a 90% efficient driver, that drops to only ~222 watts)


That's no different than - let's say - looking at a 400w HPS bulb, and expecting to get the stated lm/w from the manufacturer, when in fact you'd have to account for ballast losses (~460-480w total system draw using a typical magnetic ballast) and all the rest, either.

It's always less than what's stated; as long as we understand that, no problem. The issues we have usually lie in our expectations vs. in-system performance, not the equipment itself.

----------------------

I'll note that as an industry trend, the new CREE XT-E and XB-D's now do their binning and labeling at 85°C, which is much closer to real-world, 'in-vivo' results - and is a move I applaud.

That also means that if you can get your thermal management under control, you can (for the first time) actually get better performance than what the spec sheets list for their B&L.

Hope to see more companies following this trend.


Cheers,

-TL

Originally Posted by PsytranceorgySo it sounds to me like Sing is able to simply reduce the input voltage (via priprietary method) to the individual RED/ORANGE LEDs from 2.7 to 2.6... independent of input current..

It is not possible to lower the Vf (forward voltage, or 'input voltage') of an LED by varying the input voltage. Forward voltage is a function of the materials and process used to fabricate the diode junction (LED die). Forward voltage of an LED at the current state of manufacturing is an interesting topic- running the die hot will lower the forward voltage- but running it hot typically involves driving more current through the junction, which will actually drive the forward voltage up at a faster rate than the heat, as mentioned above, will tend to drive it down.

Note that at 700mA+, it's common to have a Tj between 70-100°C--! Don't be fooled by the temp on the heatsink exterior or the fixture; junction temperatures (which dictate emitter efficiency) are always much higher.

(Right now, almost all spec sheets are rated at a consistent Tj of 25°C. You will NEVER get this low in-circuit; that's lower than most ambient air temps just in your garden! Only way you'd get even close to that is using Peltier cooling (which is massively inefficient, and uses more than twice the watts it employs for cooling) or with a water-cooled HS, like you see in some high-end gaming desktops.

Most folks forget about de-rating performance when estimating output, simply because they don't fully understand how LEDs and typical electronic/solid state circuit losses work. But unless they've got a background in it, they shouldn't be expected to.)

--------------

With typical drivers used in these units, expect no better than ~85-86% efficiency at operating load.

So, if you were using their (Epi) cool whites (not warm), and those spec sheets were actually close to real-world performance, you might expect to actually get something closer to:

200 * 0.85 * 0.85 x 84.9 = 12,268 lumens, or thereabouts.


To get ~200w at the LEDs, you'd need an initial draw from the wall of closer to ~235w.

(With a 90% efficient driver, that drops to only ~222 watts)


That's no different than - let's say - looking at a 400w HPS bulb, and expecting to get the stated lm/w from the manufacturer, when in fact you'd have to account for ballast losses (~460-480w total system draw using a typical magnetic ballast) and all the rest, either.

It's always less than what's stated; as long as we understand that, no problem. The issues we have usually lie in our expectations vs. in-system performance, not the equipment itself.

----------------------

I'll note that as an industry trend, the new CREE XT-E and XB-D's now do their binning and labeling at 85°C, which is much closer to real-world, 'in-vivo' results - and is a move I applaud.

That also means that if you can get your thermal management under control, you can (for the first time) actually get better performance than what the spec sheets list for their B&L.

This is all really good info. It is good that Cree is rating their dies at actual real world operating temps. I would say this however- it is extremely difficult to run a typical 1mm^2 power LED's above their 'real' 350mA rating. This has been a most misleading bit of 'specsmanship' by the LED makers. Using typical, good practices heatsinking, these die junctions will be running at 80 - 90C at 350mA drive. Bad, bad, bad things happen to LED dies above 90C- which is why Cree is rating at 85C- this is regardless of what the spec for the max die temp says, perhaps a temp of 145C, something like that. That is BS, and if you talk to the Cree or Bridgelux or anybody elses' engineers, they will tell you the same thing- it's the marketing knuckleheads, not the engineers, saying this crap.

Running these dies at 700mA or more , or actually anything above 350mA, is tricky business- if it was as easy simply cranking more current through the junction to get more light then life would be easy, unfortunately real world engineering is not. Extracting the heat at these drive levels becomes a big design challenge- thermal management is everything with semiconductor light sources. You are not going there with a flat aluminum sheet as on some of the LED makers light designs, at least not with an American engineer- in China it is OK. Better watch what I say- I am in China now, Chairman Mao maybe come and get me in my sleep.

^^^^^ like

I'm beginning to believe that 90% of all panels are not capable of lasting even three years with a reasonable lumen depreciation.......

ahahhaha that might be well true!