My first dual CXM-32(170W-300W) build, for a 2'*2' space, advice about reflectors/lense etc.

[Planet Tomato]

Hello.

I am fairly knowledgeable when it comes to electronics and putting together a LED grow light is a simple matter as far as the electronics goes, but LED's is one of the subjects within electronics that I know nothing about when it comes to the use of LED's for illumination or grow lights. I am also not that knowledgable when it comes to growing but to the best of my knowlagde two 3500k CXM-32 seemed as a good choice.

I chose to use a Constant Current LED driver providing 57V - 114V(open circuit voltage max +118V) @ 1400mA - 2800mA. Rated power: 319,2W.
I am somewhat concerned about the voltage but it should be just fine, the LED's are CXM-32-35-80-54-AC00(I added a 3A fuse in series with the LED's).

The heat sink I use is a re-used 10'' x 5'' x 2.5'' finned aluminium heat sink(by the way, what is the deal with the low availability/high cost for large heat sinks? they cost a fortune compared to smaller sizes) and I hope that I will be able to runt this LED setup without a fan while it's on the lowest setting which should result in around 170W, but I am going to test out some different fans I got to see what would be best and least noisy in order to allow the full 300W.
I found this forum after having bought the LED's and driver so I am stuck with this option, since finding this forum I have learned that I should probably have split the power between more diodes but I'm pretty confident that this will work out fine.

This far I have managed just fine I think but now comes a few details which I don't know what to do about.
My grow space is very small but I don't know what that means in relation to the LED's 120deg viewing angle, if I added a lens in order to make the beam more narrow, would that increase the intensity of the light and penetrate further into the canopy? I'm simply guessing but I think that sounds reasonable enough...

What do you think, do I need a reflector/lens?

I want to keep things simple and ideally that means no lens, but if there are any real benefit to gain by adding a lens then that is what I want to do.

A lens would kind of sort out another problem I was thinking about, those LED's will probably emit a lot of heat that I don't want my plant to receive. Would it be possible to use a glass plate(maybe 3mm) or a 3mm or thicker acrylic sheet below the LED's to act as a heat shield?

I contacted the manufacturer and according to them adding a 3mm acrylic sheet would not have any optical effects while it might reduce the light intensity by perhaps 3%-5% which I don't care about if those 3-5% would create less heat directly under the LED panel.
But it might also be the case that the heat will not be a problem at the distance to the plants I will be using the light with, that is to be found out but I would not complain about if I could cover the LED's with some sort of protective barrier so they can't be touched or take damage while storing/moving the hole build.

Any and all ideas/opinions are welcomed and I would really appreciate some input on this because I am truly lost.
Warm regards

 

[CobKits]

welcome mr. tomato!

but to the best of my knowlagde two 3500k CXM-32 seemed as a good choice.

they are very efficient but not the most cost effective. the 38mm package is somewhat common but not as common as a 28mm package

id have to look at the numbers but off the top of my head 4 cxm22 should outperform 2 cxm32 at the same chip/holder cost and offer better coverage

The heat sink I use is a re-used 10'' x 5'' x 2.5'' finned aluminium heat sink(by the way, what is the deal with the low availability/high cost for large heat sinks? they cost a fortune compared to smaller sizes)

mechatronics iceled utlras are the best for cooling 38mm chips in the 100-500W range.

and I hope that I will be able to runt this LED setup without a fan while it's on the lowest setting

highly unlikely at 170W

I found this forum after having bought the LED's and driver so I am stuck with this option, since finding this forum I have learned that I should probably have split the power between more diodes but I'm pretty confident that this will work out fine.

ignore my comment above about 28mm chips being a better choice! if you have the cxm32s already we can work with that

My grow space is very small but I don't know what that means in relation to the LED's 120deg viewing angle, if I added a lens in order to make the beam more narrow, would that increase the intensity of the light and penetrate further into the canopy?

it would increase intensity... in the center. at the expense of intensity off axis. you cant make more light and in general concentrating with lenses is lossy and creates hot spots. myself and many others dont use lenses reflectors as the 120 degreeish throw works great as is

A lens would kind of sort out another problem I was thinking about, those LED's will probably emit a lot of heat that I don't want my plant to receive.

but you do...

a lens will eat up light but really wont protect your plants from heat.. if anything it will raise chip temp and reduce output and longevity. considering leds have negligible IR you wont hae temp problems as you need 5-10F higher air temps with LEDs to have same leaf temp and metabolism/ nute uptake. people talk about changing your environment with LEDs yes that is very real. i just had a horrible 2 mos because i was growing less and my overall space was a lot cooler. winter temps 70 in the days and cal and mag deficiencies under med-high veg ppfd, the girls just ate themselves and im weeks behind. i was busy and lazy and not watching them. under HID i would have had a lot more leeway,

But it might also be the case that the heat will not be a problem at the distance to the plants I will be using the light with, that is to be found out but I would not complain about if I could cover the LED's with some sort of protective barrier so they can't be touched or take damage while storing/moving the hole build.

eh they are tough. dont go spraying the emitters directly with neem oil, high EC foliars etc and you'll be fine

 

[CobKits]

rereading i see you are 170-300 TOTAL watts, not per chip

85W per chip passive is possible.. a fan wouldnt hurt

for a 2x2 you literally only need 150W so if you havent bought your heatsinks already the passives i have at my site can handle 75W without a fan, or they also can mount a small 50mm fan on top if you want the extra cooling. they are drilled for BJB holders so no soldering

 

[skoomd]

What do you think, do I need a reflector/lens?

Not in a tent.

 

[OLD MOTHER SATIVA]

took all my lenses/ reflectors off...i love cobs now

 

[Planet Tomato]

Well that sounds really good, feels good to be able to let all that lens/reflector stuff go from my mind.
Thank you very much for all the answers.

I know that 300W might be way overkill but I wanted to give my self room to perhaps later on experiment with added CO2 in which case as I understand it more light is needed.

As soon as I get some $$$ I will buy a dual or quad channel thermometer using thermocouples to attach to the heat sink to see what it can manage and not.

When I calculated the thermal stuff I assumed a operating junction temperature of 185F, but is there good reasons for why I would want a much lower temperature than 185F?

 

[nfhiggs]

is there good reasons for why I would want a much lower temperature than 185F?

Yes, lifespan is increased and efficiency is increased.

 

[Planet Tomato]

By the way, I have been thinking about adding 1 or 2 small IR LED's as well as 1 or 2 small UV LED's and drive them with very little power such that it adds something like below 1% of the total light. I can't seem to recall the exact numbers but I read a document where a LED manufacturer did a test to see if they could out perform a HPS bulb and they matched the spectrum pretty much spot on except for the lack of IR and UV which constituted a tiny fraction of the total light but some people claim that UV is important as a pest control mechanism(one reason but there are probably more), I can't recall what exactly the IR was suppose to do but it feels to me as those wavelengths might be good to have.

However I am not willing to spend much money on adding that, but a little.

But do you have any opinion about whether or not there is any point in adding UV and IR?

My LED's power are controlled by the built in potentiometer in my LED power supply so it would be hard to make the UV and IR LED's to be adjusted together with the 2 COB's but my thought was to simply add a certain amount of UV and IR and let them run on the same power no matter what the other LED's are set for.

 

[Planet Tomato]

Okey so the IR is for flowering initiation, I found out there are at least one long good thread here about that so I guess I should have researched that more.

But the issue is still the question of how to best implement this addition of IR and UV?

I am hoping to get away with using some of all my spare and scavenged power supplies to drive them, I guess the real question is how much of these wavelength would one want, estimating the light output produced for a specific power level should be pretty straight forward from the datasheet once I have selected which LED's to use.

 

[thetr33man]

Where did u get the cxm? I havent seen the larger size aground much. For 100w passive check out the
MODULED GIGA 152100

 

[OLD MOTHER SATIVA]

Okey so the IR is for flowering initiation, I found out there are at least one long good thread here about that so I guess I should have researched that more.

But the issue is still the question of how to best implement this addition of IR and UV?

I am hoping to get away with using some of all my spare and scavenged power supplies to drive them, I guess the real question is how much of these wavelength would one want, estimating the light output produced for a specific power level should be pretty straight forward from the datasheet once I have selected which LED's to use.

after three years of this led panel/cob stuff
i believe[ah belleeeeve]
you
1.need none of those initiators
2. white spectrum makes little difference 3-5k
3. same with cri
***4.its more abou thow you use bit and your grow talent and environment***

but hard driven cobs need air flow...yay fans

 

[Planet Tomato]

I got the CXM from Mouser, they stock at least two of the CXM-32 right now but I know absolutely nothing about what determines what products these site stock:
CXM-32-65-80-54-AC00-F2-5
CXM-32-50-80-54-AC00-F2-5

I can't post links due to the lack of Likes.

Those heat sinks look nice, will research them some more but they appear to be a good option.

 

[Planet Tomato]

I have now tried out my panel for a few days and on the lowest setting which should be around 170W the heatsink gets up to the following temperatures(top and middle positions refer to the position vertically in the allowable space):

Ambient temperature 77 °F.
Top position(right under the carbon filter), with no circulation fans = 172 °F
Top position, with circulation fans running on low speed setting = 163 - 149 °F
Middle position, with circulation fans on highest speed setting = 114 °F
(These values are merely approximations)

Which is not as good as I had hoped, I am really quite surprised about the 114 °F though.

But I have yet to mount fans to the heat sink, it is not a heat sink which is easy to mount fans to so I am considering simply taking some two-part glue or epoxy and sticking them on there... Not the best solution but a quick and viable solution unless there are some reason for why I don't want to use such products within the grow space, but as far as I know once the glue or epoxy has cured then there should be no problems.

 

[Planet Tomato]

I am now in the process of building a fan speed controller, first I thought about using a microcontroller to do that but in the end the fastest easiest way was to use a MAX31740.
Which is an "Ultra-Simple Fan-Speed Controller" which characteristics is all determined by resistors and capacitors and it controls the fan-speed using PWM and a NTC resistor, and I'll use that to control 2pc 4-wire fans so that I can use a PWM frequency of 25kHz which is far above our ears range of perception. I need to determine at which temperature the fan is going to run at 0% duty cycle(D) and at which temperature the fan should run at 100% duty cycle.

The temperature is sensed with a NTC resistor as the low-side resistor in a voltage divider and the high-side resistor in that divider determines the Tlow(temperature at which the fans is set to D = 0% or if I want a minimum value of D above 0) so if the NTC has a value of 10kΩ at 77°F and I use a 10kΩ resistor for the other resistor in the divider then 77°F will be the starting temperature or Tlow.

Then another resistor is set which controls the gain of an internal amplifier and long story short the gain controls the slope of the temperature/fan-speed characteristics(internal amplifier feedback resistor is 25kΩ, so a external 25kΩ resistor would equal a gain of 1, might not make sense without reading the datasheet but anyway) and what I want to ask you is this:

I have made these choices in order to create a grow light with fans that run as slow and quite as possible, but for reliability reasons, what should the maximum temperature be before the fans max out at 100%?

185°F?, 160°F?, 122°F?
or more like 105°F?

What do you think?

 

[Planet Tomato]

The obvious answer might be "try out different values and see what's fits best".