Considering a DIY LED build For flowering only: Design considerations...

turnip brain

turnip brain

Active Member
I have spent the last week reading literally thousands of posts here and on another forum about DIY Led building and design. I feel pretty good about the background info, have a pretty good basic understanding plant spectrum needs, and understand basic driver loading/ LED/driver matching. I also have the hands on shop and basic circuit building skills to be successful.

I don't assume to have the expertise to fully optimize all relevant design criteria, so I'd like to get some input.

Reasons for wanting to change from HPS to LED:
-Heat management, the 400W HPS is in a SS2 hood with high CFM cooling. Nonetheless, IR is heating up that small space way too much. With another fresh air fan circuit when doors are closed, temp in the space exceeds 95F. Bigger fans are not a solution. More fresh air flow at our super low ambient humidity will stress plants.
-less energy consumption,
-more even light distribution in the grow space, and resulting more even canopy. The HPS in a SS2 hood has very uneven distribution.

Bottom line, the space is too small to really accommodate this arrangement.


Here are my known criteria so far:

The horizontal grow area is 24" x 42" or 7 sq ft.

Height is 55" including space for the light itself. I am currently using mainlining/lst techniques to keep canopy low and even.

This build is for a dedicated light for flowering only. My vegging setup is fine for now, but the flowering setup is in need of help. There is no need for switching strings on or off to change spectrum, dimmers etc. KISS!


So, basic design concerns:

Even light distribution and relative heat sinking: building on a single large heatsink is an attractive idea in terms of construction, but perhaps building several light bars to spread out would provide better spread. Trade offs: cost for different heat sink arrangements and relative work to construct. I don't see any trouble having several light bars connected together by cross bracing into one effective panel. Any thoughts here in terms of light bars vs a massive panel?

Wattage: the guidelines i have come away with so far are for 30w/sq ft minimum, up to around 65w/sq ft. I assume this is operational wattage not combined rated wattage of LED drivers. This would mean 210-455 watts. For affordability I think i want to shoot for around 250W (I know wattage per se is not the most effective representation of useful plant light, but need a ball park starting figure) Am I in the ball park?

Emitter selection: Seems to me using Cree chips, predominately WW, perhaps with some NW mixed in would be a good basic blend/starting point for a flowering light. I'd like to nail down ratios, explore best bang for the buck in terms of chip wattage vs work of mounting/soldering more lower wattage chips. Rrog's recent build with Cree xm-l chips seems a potentially good balance in terms of using higher wattage chips so less pieces for combined output. I plan to purchase chips already mounted on stars.

Driver selection: I get the driver loading requirements/calcs. Straightforward. Just a matter of finding the best deal on the appropriate drivers when the time comes.

All input appreciated!!!
 
stardustsailor

stardustsailor

Well-Known Member
Here 's some of the BS I 've been sittin' and posting ...
My contribution ...

https://www.rollitup.org/led-other-lighting/619348-led-diy-making-dj-set.html

https://www.rollitup.org/led-other-lighting/629388-grow-project-sun_burns-every-summer.html

https://www.rollitup.org/led-other-lighting/644505-led-panel-tech-cabriolet-no.html

https://www.rollitup.org/led-other-lighting/613358-new-experimental-advanced-diy-designs.html

Caution : Research has showed that readin' them ,treats insomnia.Also they've showed a heavily 'narcotic ' effect ..
So ,better ,not to keep any glasses with beer or cups of coffee in your desk ...For your head's shake ,mostly ...
You never know ...

(Many 've fallen asleep....)

Also keep the monitor at a safe distance,from you
...Keyboard is soft with head impact and good to keep asleep onto,afterwards...
....Drooling does not make keys stick ....
Face skin ,needs ~ 30 min ,to loose ,the geometric ..patterns ,next morning ....

Enjoy ....
 
stardustsailor

stardustsailor

Well-Known Member
Additional notes ...
=> Means "which leads to " ...

explore best bang for the buck in terms of chip wattage vs work of mounting/soldering more lower wattage chips. Rrog's recent build with Cree xm-l chips seems a potentially good balance in terms of using higher wattage chips so less pieces for combined output. I plan to purchase chips already mounted on stars.

=> closely packed high power leds ,mounted on stars ....=> large heatsink .Probably active cooled .( +Fan ) .

=> powering a fan / finding a large heatsink . Properly designed & of proper Al alloy .Not a profile of somekind made out of wrong alloy .

=> Alternative power source for fan ...
=> Constant current drivers DC/DC (bucks ) or AC / DC ?
=> ~250 Watts ...In a single panel ? or 'broken' into in multiple ones ?
....



Also you might want to keep an eye on this ,as they are entering into flowering by time:
https://www.rollitup.org/led-other-lighting/650151-gd-sds-short-notice-update-13.html#post9124329
 
turnip brain

turnip brain

Active Member
stardustsailor said:
Additional notes ...
=> Means "which leads to " ...

explore best bang for the buck in terms of chip wattage vs work of mounting/soldering more lower wattage chips. Rrog's recent build with Cree xm-l chips seems a potentially good balance in terms of using higher wattage chips so less pieces for combined output. I plan to purchase chips already mounted on stars.

=> closely packed high power leds ,mounted on stars ....=> large heatsink .Probably active cooled .( +Fan ) .

Yes. understood, so is there any better trade off selecting lower wattage chips from the get go?

=> powering a fan / finding a large heatsink . Properly designed & of proper Al alloy .Not a profile of somekind made out of wrong alloy .

Yes, Planning on using an engineered heat sink with known dissipation

=> Alternative power source for fan ...

Yep, No problem there.


=> Constant current drivers DC/DC (bucks ) or AC / DC ?

At this point, planning AC/DC. House wiring supports 110-120vac

=> ~250 Watts ...In a single panel ? or 'broken' into in multiple ones ?

That's the question. I possibly envision several parallel bars rather than panels, but see a potential cooling issue trying to direct fans at long heat sink bars.


Oh, and is 250w good starting ballpark figure for 7 sq ft/relpacement for 400w hps? ....
Click to expand...

Red= replies within quote
 
turnip brain

turnip brain

Active Member
Re: previous discussion on SDS' thread my post #64: https://www.rollitup.org/led-other-lighting/658111-what-kind-light-mj-likes-7.html

Edited pertinent posts/replies:

turnip brain said:
My own application is a dedicated flowering light. I averaged SDS's flowering stages and came up with:

Blue Range 400-499nm .Peaks pref. @ ~ 410 nm , ~430 nm & ~453 nm : 8 %
Green Range 500-550 nm .Peaks pref. @ ~ 550 nm : 10 %
Yellow-Amber Range 551-599 nm .Peaks pref. @ ~ 599 nm : 10%
Red Range 600-639 nm .Peaks pref. @ ~ 625 nm : 30 %
Deep Red Range 640-670 nm .Peaks pref. @ ~642 nm & 662 nm : 40 %
Far Red Range 671-750 nm .Peaks pref. @ ~680 nm ,700 nm & ~730 nmm : 4%

(adds up to 102%, but close enough)

Then looking at eraserhead's spread, I knock off the top and bottom > 1% ranges, and combine 400-500nm to approximate SDS' 6 nm subdivisions, I come up with this:

400-499nm: 15%
500-550nm: 20%
550-600nm: 15%
600-650nm: 27%
650-700nm: 20%
700-750nm: 2%

Somewhat but not vastly different.

Average the two, roughly

400-499nm: 12%
500-550nm: 15%
550-600nm: 13%
600-650nm: 29%
650-700nm: 30%
700-750nm: 3%

So seems that for a flowering light, strain dependent, this pretty well defines a preferred spectral range.

My own challenge now is to optimize LED selection to provide a good all purpose spread in this flowering range. I don't understand yet how to read light spectrum data and graphs for specific drivers, and from reading hundreds of other posts it seems even the manufacturer's data is not a consistent comparison across the board. So this leaves me wondering how to choose specific drivers to achieve this kind of spectra for my own build.

My specific question here is, should such a panel be comprised of
1-a half dozen different specific wavelength drivers as charted
or
2-do the CW-NW-WW drivers provide enough overlap in these varying colors
or
3-would a predominately white CW-NW-WW combination supplemented with a couple other specific colors be more optimal.

Obviously, in terms of actually building a panel for my small application t would be easier to be able to select fewer variations in LEDS. Would also keep driver needs simpler.
Click to expand...
Eraserhead said:
If you're going with a dedicated flowering LED lamp, you could go a few routes, the easiest being 100% warm or neutral white, I suggest Cree for that. You could also go with either of those whites with some added reds, Cree makes a great red LED, here's the part number for the one I use XPERED-L1-0000-00801.
Kelvins are for white and nm is for colors. The lower the kelvin number, the less blue and more yellow/orange/red.
You can add colors to whites and get new correlated color temps. I use 4000-4500k (here the part # for the one I use XpGwHT-01-0000-00Hc2) and 630nm only, combined in the ratio that I use, is about 3000k.
Click to expand...
turnip brain said:
Thanks eraserhead! First question that pops into mind is; for flower then, why start at 4000-4500K rather than starting at warm white 2700 which is closer to your intended target of 3000K?
Click to expand...
Eraserhead said:
I do it for a more defined peak in the red and extra oomph in the blue, so it can be used in veg and flowering.

For example:

Veg spectrum 4000-4500k


4000-4500k + red (3000k):


And a single 2900k LED:



The levels of white stays the same from 4000-4500k, to 4000-4500k + red, it just looks lower in the chart because of the taller peak in the red.
Click to expand...

OK, so I am after a single purpose FLOWERING light. The last graph of warm white 2900K looks like it does need more red 650-700nm compared to above specific spectra averages. warm whites have enough red and far red? Is there enough blue in WW? is there enough complmentary yellow, green etc? any specific wave lengths missing or out of proportion? (I have a hard time interpreting percentage from these charts)
 
P

pepperdust

Well-Known Member
well, I am no pro..

-you want spread out lights ( more panels ).. the downfall with a single huge panel is light spread, meaning you have to pull the lamp up losing light when it reaches your plants.. see we can't compare HPS to LED, why.. HPS has a hood, and HPS shoots out in all directions, meaning the hood "evenly" distributes light over the area, and we know there are different hoods for different sizes.. small areas get small hoods, 1000's get long hoods to fill nicely.. you can't do this with a LED as of yet, unless you were to put bars of LED into a HPS hood with the LED facing up... lets just say it's a waste of light right now, maybe in 10 years.. so point being is with that much space you want a couple panels at the least, I would say 4 panels is the best configuration, unless you want to go two very long bars.. which either comes from heatsink usa, or maker led bars.

another problem I will put in right here, is there is a million combinations you can go, which really makes it hard to narrow anything down, as you need to start with some ideas ( heatsinks / wattage / chip count / drivers / any special add ons like dimmers, off switches... ) these things all make the lamp.. you could go with 10 watt chips or 2 watt chips, that makes the layout and driver change with each decision..

another thing I always have to add is fans. I think all DIY need fans regardless. heat equals less lumens, which means less efficient, I think if you build something, do it right, and add fans.. there's alot of talk of passive cooling, which just means to mean they were too lazy to add fans.. you can passive cool, your project, I will just tell you these get hotter then you think.. and I feel fans keep the temps under control giving you more lumens then if you were to passive cool.. also adding fans you can always re-use the heatsink and never need to worry once you build it, only changing chips and possibly drivers.

mounting chips to stars is easy, but if you want to buy them pre-bought that's fine. just know the stars on pre-bought are not good thermally. if you want top of the line look at sinkpad stars, dielectric stars is what you want, also a company berquiest makes them, or something to that degree of spelling. see that comes to another question, which is MONEY. how much you want to spend / can spend. DIY can be cheap as hell or expensive as hell. you need to know that whether your buying chinese drivers or nice drivers, no fans or fans, high quality fans or cheap fans.... as I said the possibilities are endless on DIY

I'm going to stop right here as I have to go, but it's good as I want you to do one thing, follow your first instincts. you can think things out forever, but trusting your intuition gets you to the finish line 100 times faster. check out some heatsinks and let us know, let us know your budget, tell us your instincts, as seriously it's as good as thinking is out. don't make the mistake and run circles forever. I feel 90% of this will be answered when you can tell us your first thoughts thinking of building a panel.. then we can help you build it and make sure it's protected from static electric discharge...... ( zeners / MOV... ) , be able to size your drivers if you don't know how and little misc. stuff


good luck, relax and enjoy. looking forward to hearing your first 2 seconds thoughts on what / how to build this. ( I'm telling you this because I didn't have his knowledge when I built mine, and trust me planning forever isn't fun.. )
 
turnip brain

turnip brain

Active Member
Pepperdust, appreciated.

I do not follow your description of light spread. maybe i am misunderstanding your point. Seems making a single huge panel with drivers distributed evenly over the entire area would be better distribution. 4 smaller panels that potentially each create more concentrated point sources would create corresponding hot spots on the canopy.

Yes, I realize the million choices.

Yes, fans are a given, not a problem

No dimmers, no switches... this was already addressed. Single purpose flowering light. KISS, no need to muck up with adjustments during the flowering cycle, just want the best average spectrum/intensity throughout flowering.

I'll consider the alternatives to star mounted chips and look into the better thermal quality stars you mention.

Quality components, not generic chinese junk. Cree LEDS as starting point as mentioned above.

I approach cost estimates from a different perspective. First explore design options for performance then narrow down to one or two basic options, price everything, then tweak for optimizing cost vs performance, then go for it. Planning isn't forever.
 
SupraSPL

SupraSPL

Well-Known Member
I agree separate modules will be the way to go for the coverage you need. It also allows for airflow around the heatsink and into the canopy.

Check out heatsink USA. You can passively cool the system all you need is the right amount of surface area. For top bins running soft I use 17in2/watt to aim for a Tj of 50c. Passive cooling has the advantage of higher efficiency (fans consume electricity), quieter and simpler design. Also the heatsink will benefit from your circulation fan in the grow room. I would be willing to bet that my passively cooled LEDs run cooler than most actively cooled designs. The downside is the initial cost, $130 for the heatsinks to run 200W of LED and that was a friend price. They are also heavy, 4 modules at 6 pounds each. Those downsides are acceptable to me but may not be for everyone.

If you are planning on running the LEDs at soft levels (700mA) and using top bins, you can replace your 400HPS with 200w of LED, coverage and yield will be no problem. I encourage 700mA as a good compromise of efficiency and upfront cost. 1A is the highest I would run an XML2 in a grow room and since XTE is much cheaper I recommend those over XML2. Also the XTE is available mounted on a 10mm star but the XML2 you would have to buy stars and reflow. Illumination supply sells XML stars for $0.50.

XML2 - 3000K - T3 bin - running at 50c
.35A 42.6%
.7A 38.7%
1A 36.4%
1.5A 32.8%
1.75A 31.6%
2A 30.1%

XTE - 3000K - R3 bin - running at 50c
.35A 41.5%
.5A 38.3%
.6A 36.8%
.7A 34.6%
1.05A 28.6%
1.5A 25.5%
 
turnip brain

turnip brain

Active Member
A little thinking out loud:

Quick relative cost comparisons for a couple hypothetical layouts using cree xm-l:

Example 1:

34 LEDs driven @ 2100ma, 3.2v ea=6.72w ea, 228.48 total watts approx $6/LED x 34= $204
2-invertronics drivers each rated to 120 watts can handle 14 chips $98 ea x 2= $196

total cost for LEDs/drivers $400

Example 2:

40 LEDs driven @ 1750ma x 3.2v=5.6w ea 224w total, approx $6/LEDx 40=$240
4- meanwell lpc-60-1750 drivers each handles 10 chips $26 ea x4=$104
total cost for LEDs/drivers $344

To me, This cost savings offsets the extra work to wire 4 strings vs 2 strings, and mounting/soldering 6 more LEDs

So opting for 40 LEDS, The question comes up, should I dedicate one string to red LEDs? Would 25% red along with WW would be excessive red? Or should I customize the fourth string with a combination of different colors, possibly including more white?


Layout for LED array:

horizontal space dimensions close to 2:1 ratio (24x42) finished panel size less than total horizontal dimensions (actually have an obstacle/post in one corner, so panel dimensions are limited to a smaller size. Practially speaking, around 36"x18" panel size would work well

Ideal layout for distribution/spread would be 4 rows of 10 or 5 rows of 8 LEDs each staggered for even distribution

As a single panel this would be a monster of a heatsink. Perhaps to optimize heatsink cost, heatsinkusa as source a 10" wide, rip into two 5" widths 36" long to have two rows of LEDs or three rows staggered per bar. So @3.15/in x36"=$113. Im afraid I don't see dissipation data for their material though.

So we are up to $344=113=$457. I have tools and solder, ac wire and plug, so no extra costs there, just need hookup wire, a switch thermal paste and screws. I am not convinced that MOV and zeners are essential, but cost for these is pretty inconsequential. I have old dead fluo light fixture as a source for sheet metal/enclosures and can fabricate something for mounting/enclosing drivers. That leaves fans and 12v power supply. All up, total around $500 plus shipping costs. what am I forgetting?

So what could i save by sourcing chips and stars and doing my own reflow mounting?

Where else could I save costs?
 
turnip brain

turnip brain

Active Member
SupraSPL said:
I agree separate modules will be the way to go for the coverage you need. It also allows for airflow around the heatsink and into the canopy.

Check out heatsink USA. You can passively cool the system all you need is the right amount of surface area. For top bins running soft I use 17in2/watt to aim for a Tj of 50c. Passive cooling has the advantage of higher efficiency (fans consume electricity), quieter and simpler design. Also the heatsink will benefit from your circulation fan in the grow room. I would be willing to bet that my passively cooled LEDs run cooler than most actively cooled designs. The downside is the initial cost, $130 for the heatsinks to run 200W of LED and that was a friend price. They are also heavy, 4 modules at 6 pounds each. Those downsides are acceptable to me but may not be for everyone.

If you are planning on running the LEDs at soft levels (700mA) and using top bins, you can replace your 400HPS with 200w of LED, coverage and yield will be no problem. I encourage 700mA as a good compromise of efficiency and upfront cost. 1A is the highest I would run an XML2 in a grow room and since XTE is much cheaper I recommend those over XML2. Also the XTE is available mounted on a 10mm star but the XML2 you would have to buy stars and reflow. Illumination supply sells XML stars for $0.50.

XML2 - 3000K - T3 bin - running at 50c
.35A 42.6%
.7A 38.7%
1A 36.4%
1.5A 32.8%
1.75A 31.6%
2A 30.1%

XTE - 3000K - R3 bin - running at 50c
.35A 41.5%
.5A 38.3%
.6A 36.8%
.7A 34.6%
1.05A 28.6%
1.5A 25.5%
Click to expand...
Doh! was writing my post when you posted this, I'll chew on this info and read your thread: https://www.rollitup.org/led-other-lighting/590648-diy-led-220w-cree-xte.html
 
P

pepperdust

Well-Known Member
turnip brain said:
Pepperdust, appreciated.

I do not follow your description of light spread. maybe i am misunderstanding your point. Seems making a single huge panel with drivers distributed evenly over the entire area would be better distribution. 4 smaller panels that potentially each create more concentrated point sources would create corresponding hot spots on the canopy.

Yes, I realize the million choices.

Yes, fans are a given, not a problem

No dimmers, no switches... this was already addressed. Single purpose flowering light. KISS, no need to muck up with adjustments during the flowering cycle, just want the best average spectrum/intensity throughout flowering.

I'll consider the alternatives to star mounted chips and look into the better thermal quality stars you mention.

Quality components, not generic chinese junk. Cree LEDS as starting point as mentioned above.

I approach cost estimates from a different perspective. First explore design options for performance then narrow down to one or two basic options, price everything, then tweak for optimizing cost vs performance, then go for it. Planning isn't forever.
Click to expand...

sorry if you don't get my light description, maybe hard to explain..

thing is, your huge panel is going to be the same size as 4 separate panels.. so if you could find a heatsink that wide ( which i don't think exists ) then by all means have at it. the lowest your gonna get by is have a 2 long heatsinks ( 6 inches wide by 36 inches long minimum ) . or you can split the panels up into smaller sizes. you could connect the panels as one huge panel with connectors, or yes you might have to have seperate drivers for each panel, as you said you say RROG's panel, look how many drivers he had for one panel..

I get back to the light again, as it's something people don't understand yet with LED. I good average is 120 degree spread. well HPS is 360 in any degree you look at it from. HPS has us beat in light spread, but with that comes in-efficent light, as it has to be reflected back before being able to be used ( top half of the bulb if you cut it across ) that is un-usable as it would go up, and then minimally get reflected once it hits your ceiling first, hence why the reflector was made... with LED that doesn't exist as most LED put out a 120 degree angle making reflectors pointless as no light is being angled upward. now you can focus LED light with reflector "caps" or concentrate it with lens, all which come at a cost of light to reflect it, even a reflector cone is like 92% efficient. which is why we came up with the idea right now of seeing "bare LED" as any beam angle you want is going to come at a cost of lumens. some seem to think even with the losses it is still better, but not all of us are sold on that yet... I myself am not a mathematician, so no help there

now take your 120 degree lens on the diode and try and reach 1 1/2 feet across to reach the the corner plant.. it's not going to work is it. well what do you do, you raise the light till the 120 degree pattern hits the corner plant. problem solved right? no, you just raised the light so high to reach the corners of your grow that you just made your light half the intensity it was.. that i no way solves the problem we want to solve as our plants eat high levels of light, and more light = more calyx.

this we can never change about a LED.. the drawback of LED and will always be light spread / amount of chips to level out light intensity levels. there is no way we know of around this yet, and I don't see the problem changing anytime soon. the most likely solution is high watt chips so we use less, but it doesn't matter if you have a 100 watt chip, it's not the same as a 100 watt HPS, and it never will be, then we would have to go different angled position LED's, along with back to what a HPS has to do, reflector, or hang it vert...

every single LED light you see that has 200+ watts in a small area is in-efficient made, it's that simple. do they work, hell yes they work, but as I said at the cost of raising the light to hit all the areas where it's covering. a truly efficient LED panel is spread out, plain and simple. I'm not saying it's ideal or good, but it is what it is, until someone comes up with something, which might be as simple as multiple bars on the unit where you can angle the bars, that could work I see. but you don't see it in any panel but one ( which is a flower booster, illumitex safari ) . unless you could find a curves piece of metal like half of a pipe, I suggest you spread out the LED, and not concentrate them.. that makes for the hotspot you mentioned.. also maybe your thinking of my version wrong, as in your mind the light should be very evenly spread with no hot spots at all ( obviously the middle gets more light by a little )
 
turnip brain

turnip brain

Active Member
pepperdust said:
sorry if you don't get my light description, maybe hard to explain..

thing is, your huge panel is going to be the same size as 4 separate panels.. so if you could find a heatsink that wide ( which i don't think exists ) then by all means have at it. the lowest your gonna get by is have a 2 long heatsinks ( 6 inches wide by 36 inches long minimum ) . or you can split the panels up into smaller sizes. you could connect the panels as one huge panel with connectors, or yes you might have to have seperate drivers for each panel, as you said you say RROG's panel, look how many drivers he had for one panel..

I get back to the light again, as it's something people don't understand yet with LED. I good average is 120 degree spread. well HPS is 360 in any degree you look at it from. HPS has us beat in light spread, but with that comes in-efficent light, as it has to be reflected back before being able to be used ( top half of the bulb if you cut it across ) that is un-usable as it would go up, and then minimally get reflected once it hits your ceiling first, hence why the reflector was made... with LED that doesn't exist as most LED put out a 120 degree angle making reflectors pointless as no light is being angled upward. now you can focus LED light with reflector "caps" or concentrate it with lens, all which come at a cost of light to reflect it, even a reflector cone is like 92% efficient. which is why we came up with the idea right now of seeing "bare LED" as any beam angle you want is going to come at a cost of lumens. some seem to think even with the losses it is still better, but not all of us are sold on that yet... I myself am not a mathematician, so no help there

now take your 120 degree lens on the diode and try and reach 1 1/2 feet across to reach the the corner plant.. it's not going to work is it. well what do you do, you raise the light till the 120 degree pattern hits the corner plant. problem solved right? no, you just raised the light so high to reach the corners of your grow that you just made your light half the intensity it was.. that i no way solves the problem we want to solve as our plants eat high levels of light, and more light = more calyx.

this we can never change about a LED.. the drawback of LED and will always be light spread / amount of chips to level out light intensity levels. there is no way we know of around this yet, and I don't see the problem changing anytime soon. the most likely solution is high watt chips so we use less, but it doesn't matter if you have a 100 watt chip, it's not the same as a 100 watt HPS, and it never will be, then we would have to go different angled position LED's, along with back to what a HPS has to do, reflector, or hang it vert...

every single LED light you see that has 200+ watts in a small area is in-efficient made, it's that simple. do they work, hell yes they work, but as I said at the cost of raising the light to hit all the areas where it's covering. a truly efficient LED panel is spread out, plain and simple. I'm not saying it's ideal or good, but it is what it is, until someone comes up with something, which might be as simple as multiple bars on the unit where you can angle the bars, that could work I see. but you don't see it in any panel but one ( which is a flower booster, illumitex safari ) . unless you could find a curves piece of metal like half of a pipe, I suggest you spread out the LED, and not concentrate them.. that makes for the hotspot you mentioned.. also maybe your thinking of my version wrong, as in your mind the light should be very evenly spread with no hot spots at all ( obviously the middle gets more light by a little )
Click to expand...
Thanks for the long explanation. Everything understood. I indeed did misunderstand your initial post. I think we are more on the same page than not. Even distribution of LEDs across area is exactly what we agree upon! The idea of angled bars is new. Hadn't heard that one before.
 
P

pepperdust

Well-Known Member
^ glad your on the same page..

second, I have to correct my first statement, even a same size heatsink as smaller ones is still not efficient as smaller ones, for the fact the smaller ones have gaps, where the big heatsink wouldn't have any gap.. that's what makes the smaller ones more efficient, is allowing the spread out, so to make my statement correct, you would have to have 1 1/2 wide to make it as efficient, then you would have more heatsink for heat management.. but I'm going to stop there.. I think you get the idea.


we are in the second stages of LED , to me first was red and blue, now were adding more watts, and different colors, what's next who knows. but LED is not for everyone, there is still alot that is going to change that is guaranteed, if you read into LED news is mind boggling what there coming up with.. were just starting, but it looks to be more efficient then HPS if done correctly, and hitting GPW very very easily over a HPS.. the future is going to be awesome... jumping on board now I think is really only to get accustomed to it, like growing your first time is to now growing.. not that there isn't benefits right now to LED, but as I said we are in the baby stages of LED and growing pot..once we figure actual spectrum, and LED's advance in lumens far over what they are, and new technologies come into market, HPS is going to look like what a incandescent looks to T5's, when comparing HPS to LED
 
turnip brain

turnip brain

Active Member
Quite a romantic (and pretty realistic) outlook.

Now is good for me! To overcome the stated drawbacks of my HPS now is a good thing.

Starting with good heat sink seems the base for ongoing upgrades as emitter technology evolves.

But who knows what the next technology will be?
 
guod

guod

Well-Known Member
Im afraid I don't see dissipation data for their material though.

hidden in the text...

Quality extruded aluminum heatsinkWidth is 10.000"Fin Height is 1"Base Height is .300"Weight is approximately .50 lbs per inch C/W/3": approximately .90 All of our heatsink extrusions are sold by the inch. ...
 
turnip brain

turnip brain

Active Member
guod said:
Im afraid I don't see dissipation data for their material though.

hidden in the text...

Quality extruded aluminum heatsinkWidth is 10.000"Fin Height is 1"Base Height is .300"Weight is approximately .50 lbs per inch C/W/3": approximately .90 All of our heatsink extrusions are sold by the inch. ...
Click to expand...
OK, but I'm afraid I don't know how to interpret that. Please help me out.
 
turnip brain

turnip brain

Active Member
guod said:
Im afraid I don't see dissipation data for their material though.

hidden in the text...

Quality extruded aluminum heatsinkWidth is 10.000"Fin Height is 1"Base Height is .300"Weight is approximately .50 lbs per inch C/W/3": approximately .90 All of our heatsink extrusions are sold by the inch. ...
Click to expand...

I can do the math, but I don't understand what is notated here... C/W/3": approximately .90

Anyone?
 
ya bongo

ya bongo

Well-Known Member
C/W/3": approximately .90
let me translate it for you...

this Heatsink has a thermal resistance (absolute) of 0.9 C/W for a lenght of 3"

...but this is useless if they don´t give a diagram for other lengths.

btw.
I also now understand, why all here calculate heat sinks over aera... lol:wall:
 
turnip brain

turnip brain

Active Member
ya bongo said:
C/W/3": approximately .90
let me translate it for you...

this Heatsink has a thermal resistance (absolute) of 0.9 C/W for a lenght of 3"

...but this is useless if they don´t give a diagram for other lengths.

btw.
I also now understand, why all here calculate heat sinks over aera... lol:wall:
Click to expand...
Yeah.

I have a spreadsheet calculator from another forum, that calculates dissipation in watts from number of fins and all dimensional info, but don't see how C/W/3": approximately .90 applies to that. Seems like a different set of parameters and equations.
 
turnip brain

turnip brain

Active Member
OK, so back to FLOWERING spectrum, seems that part of the discussion just dropped out. Still have a couple questions

Based on averaging SDS and eraserhead's recommendations, I came up with:

400-499nm: 12%
500-550nm: 15%
550-600nm: 13%
600-650nm: 29%
650-700nm: 30%
700-750nm: 3%

The simplest arrangement seems to be all WW around 2700-3200K,

The resulting red peak is around 610nm, and there is another blue peak near 450nm.

From what I have read, seems blue is covered OK for flowering. Is that correct?,

Should I supplement with more red/far red to increase roughly the 640-700 range? If so, plugging in and looking at combined spectra here: http://buildmyled.com/custom-led-strip/, adding red/far red creates intense peaks in their own range relative to the overall curve of just WW, and change the entire scale of he graph. I wonder how much is too much additional red/ far red and how to balance relative outputs if I do add them to the WW. It is not intuitive nor do I have the experience/expertise to assess this when the addition of red creates such profound changes in the spectral graph.

Great appreciation for educated and especially empirically reinforced input!!! I am no dunce :dunce:, but sorting so much conflicting info across the web leaves me feeling like one! :wall:
 
You must log in or register to reply here.
Top