A customised light design - some input needed.

[koosjr]

As an engineer, I cannot help but to design my stuff as well as I can before I build.

I would appreciate some input on the concept below. It is still early and good ideas can be incorporated.
 

I am in the process to plan my own room. However, I am off grid so power input to the system is a big concern for me. I need to get very economical on power use when I have no solar.

I understand that cannabis are a bit different than other plants and prefers blue spectrum light during the vegetation phase and red spectrum during the flowering phase.

Currently, the information I have is that the following LUX levels are required:

 

Seedling: 7000 Lux

Vegetation Stage: 15 000 - 50 000 Lux

Flowering Stage: 50 000 - 70 000 Lux

 

I am currently busy with a LED light design, and I can find fairly good LED's with an efficiency of about 120 lm/w for both 3000K and 5000K.

The light will have 4 channels: 1/3 5000K, 2/3 3000K, a bit of IR at 840nm and I consider some UVA at 365nm.

That said, the UVA LEDS are seriously expensive and I cannot find conclusive proof that it has any positive effect. However, it is easier to make provision for this in the circuit and leave out the components if one choose to.

All channels will be fully dimmable, so one can for example blend the 3000K and 5000K if you want to which should give really good control between red and blue spectrum for those who wants to experiment.

 

I want to work on a LUX level of 25 000 for the 5000K, which if one use 18/6 timing, I will have to run at least for 8-10 hours from a battery or generator. In his time I want to use as little power as possible.

For the 12/12 stage I want to work on a Lux level of 60 000 for the 3000K. In this case 4 hours will be out of solar.

In total then, IR excluded, a lux level of 85 000 would be possible if all circuits are used. Too high I know, but read further and you will see why.

So, if in the day during vegetative stage I have more solar power, I can dial in some 3000K to the 5000K to increase lux levels if I want to.

I assume IR and UV will burn all the time.

 

Right, so in terms of maintenance, I want to make the panels a bit more modular as I hate things that are not serviceable, and secondly provides no redundancy. 

My experience of the past few years in the building industry, is that not even from the premium brands, we get the quoted life they claim any more. So, I am going to assume that some circuits will fail prematurely.

For the reason above, I decided to work on an external 27 VDC supply. Although this makes wiring more cumbersome, the LEDS can be arranged such that should any fail, at most a string a of 8 LEDS will go dead. This will at least provide some form of redundancy on the same fitting.

I can go 24VDC too, but that will require a 7-string LED. It is quite a challenge to match up the different voltages as some of the LEDS rapidly loose performance with only a 0.1V difference on the input.

I also like low voltage in the room more than high voltage especially since one can have problem with moisture in the room.

Also, with external supplies, it can be placed outside the room so that the heat load to the room can be smaller if required.

The modules will be repairable if you have a SMD station so individual LEDS that failed can be replaced.

In terms of control, I have not quite decided yet what I want to do. I can do an external controller and daisy chain a bus cable between the lights. If the controller fails in this case, one can at least switch the lights manually with a simple work around.

I can also use an ESP32 and make the control wireless. If the software fails, you may need to go to each light individually to switch it.

What I do want if there is a UVA channel, is to have a door switch that will automatically switch it off if the door opens.

 

Currently, I do want to save some costs, so I do consider the following:

1. Treat the PCB with a conformal coating to make it water proof and have no enclosure. I will be suspended as a bare flat panel - another reason for going 27VDC

2. I the above case, all I do then is to simply increase the size of the PCB. The LEDS use the copper layer as cooling and with enough surface area there one would not need any fans.

3. A further advantage then of a larger PCB is obviously that the lights are further apart which will help a little with better light distribution.

4. The board will be designed to run at 60°C at an ambient of 35°C. I can for argument sake also put a temperature sensor on the board to monitor the light.

There is really a lot of things one can do. Does the above sound feasible for those who want more control of their lighting?

Edited May 14, 2023 by koosjr

[Ill_Evan]

Sounds pretty on-point to me my dude. I wish I had known how big you wanted to go before answering in your other thread 

My only real pointers is that more often than not it is more important to focus on environmental factors for optimal growth than it is for how much LUX your are working with. LUX is cool but most people here will use umol when deciding on light power. The standard at the moment is to achieve 700 umol at max during flowering stage. Any more than that is useless as the plant can only physically make use of 700 umol without carbon dioxide injection into the room. 

Some growers will find that their plants stress when running lights at maximum or that the environment gets too hot and stunts the plant. This is why I say environment is more important than lighting. Yes you need good lighting, but it isn't the only important thing when growing cannabis. 

2 hours ago, koosjr said:

That said, the UVA LEDS are seriously expensive and I cannot find conclusive proof that it has any positive effect. However, it is easier to make provision for this in the circuit and leave out the components if one choose to.

Definitely right on there. UVA LEDS in my opinion are gimmicks. 3000k or 3500k full spectrum white light LEDs are all more than perfect to grow top tier bush. 

Interesting that you want to push 5000k during flowering. If you wanted to do that, I'd maybe suggest combining both the 3000k with the 5000k at the same time. Blue light helps keep the plant nice and short. The more red/yellow the light, the more susceptible you may be to over stretching. This will be strain specific though. The 5000k should in theory help push flowers more efficiently than the 3000k, but I still think you need some of the 3000k spectrum at the very least during pre-flower and then halfway through flower push just the 5000k. I'd personally be interested in the results.

Just for reference, I have done full grows with both just the 3000k and 3500k with the same strains and in all honesty, I haven't noticed much difference in quality but the plants growing with the 3500k did stretch more during growing. 

[koosjr]

HI III_Evan,

 

I am for no forced to work on Lux values as I have no other data on the LEDS in the sheets, so that is the reference I have.

I made a Typo. I wanted 60 000 Lux on 3000K.

But you are now making new suggestions.

I thought that 3000K was almost exclusively needed for flowering stage, but you now say I should start with 3000K and later move over to 5000K.

I assume though that the IR will work all the time.

Do you mean that I should make provision then for 60 000 lux at 5000K?

This sort of means that I will start vegetation stage at a 50% load on the 5000K and eventually add some 3000K.

I will then have to change over to 3000K for flowering - perhaps with some 5000K in there and end of with only 5000K?

If this is the case, then I will have to double up on my 5000K portion.

I did investigate using dedicated 660nm LEDS but they are so low in power output that they will not add anything useful.

Edited May 14, 2023 by koosjr

[Ill_Evan]

I apologies, I got the kelvin colours backwards. I meant the opposite, 5000k for veg, 3000k for flower.

The principle though is the same, start with 5000k and then add the 3000k later in flower. You'll also be quite surprised at how little light is actually required for veg stage. For a 1.2mx1.2m tent I can easily veg with just 240W. 

[koosjr]

Cool, then we are on the same page.

 

That is then why I have less 5000K in my lamp.

Right, so I should still add some 5000K during flowering to keep the plant shorter then which in the end can be switched off completely.

If you say 240W on 1.2x1.2 that basically speaks to my calculations - that is now of course if we assume the LEDS are similarly efficient.

I see that only the very white LEDS upwards of 6500K achieve up to 200 lm/w values, but they offer very limited performance at 660nm and even 550nm- which I think is rather important.

 

BTW, I have seen a light here done by a member where he have put together a number of single row strips. That is also something to consider because one can then insert small amounts of 6500K lights if you want play around a little.. Engineering. Always a balancing of compromises.

 

That said. Seems I better change to 24V as this is by far a more common power supply than the 27V. The latter however is perfect to keep a 24V battery on standby.

I will be loads cheaper than a Samsung light for R 21 per watt.

[Ill_Evan]

Always cheaper to build than to buy. The prices are slowly coming down and prices from Shenzhen directly are also not terrible. But definitely build it yourself if you can, Mr. Manly Engineer man 

If you can hook up a potentiometer and a small watt display that'd also be tops. 

IR in my opinion is also a bit of a gimmick such as UVA. Again, a lot of people start throwing in all these wavelengths and lose focus on the environment. Those extra chips won't do anything for a fungus outbreak or a bad pH. It's all about the umol at the end of the day. 

For a standard 240W QB panel you're looking at a umol of roughly the below, just as reference. Remember we want about 700 umol max during peak flower. It sounds like your lux targets should be correct. I run my boards at 30cm from the canopy at all times from the point that the plant is bigger than 30cm itself. At max per 1.2m² during peak flower I'm running about 440W. At the moment that accommodates a max of 5 plants, and I am experimenting with running 4 or 5 at the moment with different veg times and training to see what is most efficient. 

[koosjr]

OK, so the umol values are probably not related to lux as the lux meters are tuned to measure human perception.

Do you get a specific meter or sensor for that?

[Ill_Evan]

1 hour ago, koosjr said:

OK, so the umol values are probably not related to lux as the lux meters are tuned to measure human perception.

Do you get a specific meter or sensor for that?

Apogee makes a par meter, locally it would be something like this. 

[koosjr]

Ouch! I can see why that sensor is so expensive.

I don't know of a single component that has that range. It probably has multiple elements and some fancy software inside.

[Ill_Evan]

They were like R12k easy when they first came to the country. There were a few guys that would lease them out a day at a time so guys could just calibrate their setups. I'm sure there are still a few grow shops that would do it, or at least know of someone with it. 

[koosjr]

https://migrolight.com/blogs/grow-light-news/cheap-par-meter-hack-use-a-lux-meter-to-measure-par-accurately
 

Perhaps not a bad ay of doing it initially.

[Ill_Evan]

3 hours ago, koosjr said:

https://migrolight.com/blogs/grow-light-news/cheap-par-meter-hack-use-a-lux-meter-to-measure-par-accurately
 

Perhaps not a bad ay of doing it initially.

I'm aware of MiGro, clever dude. You can definitely do it that way as well. 

[koosjr]

Question:

What if I incorporate the light fitting and the ventilation?

Make it a low density 400x400 light fitting with perforations through which one can then also have a very nice downward flow through the plants?

4 fittings on a 1.2x1.2m tent.

[koosjr]

OK, so I found that LED light efficiency is about 40%, which means that in this room I will have a heat gain of nearly 300W.

So, if I do blow air through my light fitting it means at least a 3°C increase if the air passing through the fitting. That will give me problems in summer but in winter this will be advantageous.

I need to think a little about it...

I won't have enough power to use an AC unit, but I do have other options. Maybe some geothermal...

[koosjr]

Nope, ventilation for the room through the light fitting will probably not work as well as I hoped.

I ill have to get that heat out separately which means extraction and I doubt that ill give much leave movement.

Some good ideas on this forum help shaping my ideas.

Edited May 15, 2023 by koosjr

[DesignatedDave]

16 hours ago, koosjr said:

Ouch! I can see why that sensor is so expensive.

I don't know of a single component that has that range. It probably has multiple elements and some fancy software inside.

Download the Photone app for your phone and use white paper and tape for the diffuser.

This measures PPFD, LUX, DLI, etc and its pretty damn accurate.

 

 

[Naughty.Psychonaut]

Hi bud, hope you well. We got a inhouse LED specialist, if I may say so myself. @MrE have you seen this? 

@koosjr I did a quick over everything, but I didn't pick up on any mention bout if you gona be using the light for veg or for flower or for both? 

when buying a light this most important so you can imagine how important it is to know what you wana do with the thing before diving head first.

a duel spec light that will cover both veg and flower phases needs to be quite a bit stronger in overall wattage and you'll need to be able to switch between spectrums like turn the red on and off by itself, same with the blue and you gona have to mix spectrums to keep the plant happy through all of it's cycles. Duel spectrum lights don't really offer that. If you take a 300w driver and make it pump a duel spectrum - even if it can't switch between spectrums and just on full blast all the time - it means during veg your plant is not getting optimal PAR/PPFD readings and undesired spectrum for vegging, and same goes for flower. not technically but to make it easier to understand, then the plant will only getting 150w of good veg light and 150w of good flower light, cause you pulling that 300w in two different directions. It's more technical than that, but I am sure you get what I mean? So building a duel spec you need to get a driver that can push 600w and split that spectrum so you can toggle between veg and bloom.

basically what I am saying, it's the same as with all things in life, if you want one gadget that can do multiple things it becomes gimmicky. it will do both things at mid quality. 

much better to have, lets say 300w on full blast, but with the right spectrum in order to get an optimal veg and a optimal flower light. 

you'll be glad to find out that those added UV, added IR and all the mixing and splitting of spectrums you only need to do if you want to go for a duel spec light. which I would advise against. much more important to be able to dim the light, when it's in the right spectrum the output % becomes way more important.

Besides what we telling you, we can talk till we blue in the face, but here is a link to LED gardener, I'll drop the link to the page you need to be on right now study it thoroughly, there are even step by step guides for each build, you get optimal performance minimal mumbo jumbo. 

https://ledgardener.com/diy-led-strip-build-designs-samsung-bridgelux/

 

[koosjr]

I think my first post is clear enough what I want to achieve.

Yes, I am going to have a light with both 3000K and 5000K such that either can provide the full power needed for veg and flower.

I will then have each circuit to be fully modulating - which is easy enough to do so I can go 50% of the one and 50% of the other if I want to.

It seems to save manufacturing costs, I need to limit my panel sizes to 400x100mm - not a bad length and I am still playing with ideas of how to assemble it.

I want use aluminium PCB's as it will help a lot to remove heat from the LEDS and hopefully will at least make them last better.

My thinking starts leaning towards having a dedicated plenum system for the lights where the air in/out is a separate circuit to that of the airflow for the box - so this means that almost the entire heat load of the lights can be removed from the box.

[Naughty.Psychonaut]

31 minutes ago, koosjr said:

I think my first post is clear enough what I want to achieve.

Yes, I am going to have a light with both 3000K and 5000K such that either can provide the full power needed for veg and flower.

I will then have each circuit to be fully modulating - which is easy enough to do so I can go 50% of the one and 50% of the other if I want to.

It seems to save manufacturing costs, I need to limit my panel sizes to 400x100mm - not a bad length and I am still playing with ideas of how to assemble it.

I want use aluminium PCB's as it will help a lot to remove heat from the LEDS and hopefully will at least make them last better.

My thinking starts leaning towards having a dedicated plenum system for the lights where the air in/out is a separate circuit to that of the airflow for the box - so this means that almost the entire heat load of the lights can be removed from the box.

just shows you how quick my quick over was, but as soon as I read you mixing spectrums and adding uv and ir and that stuff then I knew where it's going. 

if looking at doing a duel spec, working with the 300w/1m2 rule, you gona have to put roughly 600w driver on. cause you only using half the light most of the time if I am not mistaken? you going for building 2 different lights on one panel? gona be a tough one but goodluck!!!

have you seen the LED BAR designs??? no need for extra heat management and stuff brother. You already got the PCB's? I'd advise to go with the aluminium extrustions. look for a T shape. Pack of tiny pop rivits, tiny drill bit. Build a square with bars running across. light will basically never even heat up + coverage and penetration goes waaaaay up since you have diodes all over the canopy and you can bring the light muuuuuuch closer to the canopy cause of the little heat. just all around better design. don't have to run 100% all the time + minimal heat = last almost forever with minimal power drop.

for ages and ages the techy techy dudes been teching the tech out these tech things, like a couple years ago you could find these big body LEDs with built in fans and and and all the fancy jazz, all the extra shit they didn't need cause they too focussed on the tech side, not looking at design. a lot of times, when we approach something we wish to understand better cause we don't completely understand it yet, it's always a complicated start so we tend to look at the easy answers and go "no way it could be this easy, it has to be more complicated" when actually it really is that simple. just the layout of the light design can make a moerse difference in the way the tech performs. with those old big body LEDs with built in shit you buy a light that sucks 300w from the wall, but 80 to 100 of those watts are being put out in the form of heat and another 50w goes to running the fans leaving you with about 200w of light output, then people are left scratching their heads when they throw a PAR meter and get less than what the product is advertised, just cause the design is all wrong.

[koosjr]

Yes, basically it is two different lights on one panel so although the total power of a panel will be say 200 W, I will be using only 100W for Veg and 200W for flower.

I surely have no intention to use both spectrums all the time.

I also want it dimmable because if I run from solar, then some days I might have to compromise and tune down the light to the minimum level possible for the hours that I have limited or no solar.

I do not want to get too technical though.

Where I don't follow you well is with regards to the heat rejection in the box. Event he 165 lm/w Samsung LEDS have almost 50% energy loss in heat. So, if you power 100W, then 50W is heat.

Do I miss something there?

 

 

[Naughty.Psychonaut]

To achieve optimal desired spectrum on the leaf surface for both veg and flower situations with one light is almost impossible, because they're so far apart. This is the reason people mix and match, but if you upping the Wattage and switching between spectrum, it's better not to play in the middle. Best to go full white at 5500k for veg and switch to full yellow at 3000k for flower and look at it as two seperate lights. much less technical and you don't have to worry bout adding blue and red and and and and and, cause that stuff just takes you closer to the middle again. and adding red to 3000k spectrum is over kill, my flower light is a DIY 3000k already can't see shit cause it's so yellow it looks like HPS. same with white, adding blue to the 5500k is overkill. it's only when you work around the 4000k mark that it really gets technical cause then you have to add blue and then you look at one single light as "best of both worlds" but it actually does two things half half and cause it's more technical it's easier to go wrong. 

But that's just my  

About that last bit, looking at the bigger picture, out of all lighting options, the life span of LEDs beat out all the rest, except for the centennial light bulb that's been burning for 120 years straight but yeah, so getting into why it lasts so much longer is because the output in light is so much greater than the output in heat, even from the little generic poop grade LEDs you get on your christmas lights. Samsung, with their new tech, they start pushing the limits. Some diodes give off more heat some give off less heat, but in all cases it's not a fixed output, it's always got to do with the resistance you put on it. You're an engineer, I just grow plants hahah you probably know waaay more than me about thermal resistance and that stuff? 

If you look at 2 of the exact same diodes, brand new. One running one at 100% output will wear it out quicker than that same one runming at 50% output. The reason for that is because the 100% diode is running physically hotter than the 50% one. and this is the same principle that goes for why most electrical appliances need fans on them to keep them running cool at all times to keep them from wearing out too fast. most electrical stuff got built in fans for that ot the design of the body allows for passive cooling of to some degree. The hotter a circuit runs the more resistance it has on it, the more resistance the harder the thing has to work to reach the same amount of output and eventually shortening the lifespan much quicker.

Logically, if the design or "layout" of your light causes it to heat up more the light will have a shorter life span. If you gotta run fans on it, it means it runs hot. If you gotta run more fans you pulling more W from the wall. why not build a better working design that the plants will also appreciate more PLUS save money by not having to run extra stuff plus less wear on the light 

[Naughty.Psychonaut]

Oh wait, maybe I am not even answering you there. just rambling away!

The bar LED layout design allows for passive cooling, so you don't have to actively cool it with extra fans. With the bar layout, out of all designs, allows for the coolest running temperatures from your diodes and driver. Allowing space between strips and driver it doesn't build up or trap any heat. you can hold your hand around the strips while on, air moves around the strip freely, extrusion strip being alu it doesn't conduct heat, light spread muuuuuuuuch better allowing for closer to canopy allowing for better penetration and overall less heat, allowing for less electricity consumption cause no extra fans

[koosjr]

OK, so I do understand better now what you say - but you are mixing up your 3000K and 5000K again hehehehe..

For this discussion, I have attached the data sheet of the LED as it will help a lot to explain a few points.

Your input is based on experience and it helps me a lot to build this puzzle quicker. Thanks a lot for that.
 

Yes, my light will indeed be such that if I use the 5000K channel it will be enough light without using any 3000K and vice versa.

Regarding the heat though, there I have to differ from you completely.

My day job is HVAC so I have been doing heat load calculations for buildings for the past 20 years.

Any device, regardless if you can "feel" the heat or not, if it dissipated say 20W of heat, it is done - it does not matter if it was done at a surface temperature of 40°C or 60°C. You can't reverse it.

Heat from a device (and even humans) is basically transferred in two ways to the room. Convention and radiation.

If that heat was transferred to the room air either way, eventually the room air heats up the same amount and you have to remove it via the AC unit or otherwise.

 

However, a higher surface temperature does make a difference with regards to how heat is transferred as the higher it is, the larger a portion of the heat will be transferred via radiation.

Radiation can be a very bad thing because that is the thing that can literally burn your plant leaves from a distance. 

Spreading out the LEDS on the panel to lower the surface temperature therefore will reduce that radiation component - but not only that, it spreads the light out better.

What you are trying to suggest is that if you run a LED at a lower output, it will be more efficient. I did not thought about making those calculations, but yes it does bring up the efficiency slightly. If I work on 90% of the rated lumens, then I go up from 137 lm/w to 151 lm/w and at 75% lumens it reach 160 lm/W.

That said, the difference in heat rejected between 137 lm/w to 160 lm/w is not that much. I can however see the advantage in longevity clearly.

What is also tells me is that controlling the LED means I basically need a current controller - not a voltage controller.

Since in a room like this, lights are the major source of heat, it can be a big advantage if one can capture that heat and get it out of the room before it is transferred to the room air.
 

All in all, we can get to a very good answer here. An affordable, reliable light that does not dump heat into the room.

2009021504_HONGLITRONIC-Hongli-Zhihui--HONGLITRONIC-HL-A-5730D1W-S1-08-HR3-LY_C210347.pdf

[Naughty.Psychonaut]

7 hours ago, koosjr said:

OK, so I do understand better now what you say - but you are mixing up your 3000K and 5000K again hehehehe..

hahah brother if you switching between 3000k flat and 5000k flat you not mixing at all. when the one is pumping other one is sleeping. never running both at same time. 

going with 4000k stuff you gotta add either red or blue - mixing spectrums. that's what I mean by that. cause already 4000k isn't where you wana be so you HAVE to add blue for veg to pull that 4000k closer to 5000k (mixing the stuff) and for flower you still work with same 4000k and add red to pull the 4000k closer to 3000k. and the 4000k will be running at all times and you only switch beteeen red spec and blue spec. that's early 2000's tech, it's been improved on. 

heheee bout that heat thing, we in the same book but just on a different page you say you don't wana get too technical, but that's waaaaayy more technical than it needs to be. 

Forget about the lm/w forget about all the finer detail stuff. Basically all I am saying is - the bar style layout design allows for the light to run cooler over all. Now you already got the common knowledge on how something that runs hot is wearing out faster so that's point nr.1 but you also saying you wana displace the heat from the light before it heats up the room? why not skip all that and go with a light design that doesn't heat up? saving you money in the long run (not having to replace burnt out / worn out parts) aswell as the short run (not having to run extra fans or cooling appliances of sorts) aswell as providing better light coverage and penetration for your plants? It's a win win win situation.

going with PCB is shooting yourself in the foot, people used to use these early 2000's but we have discovered new and better ways. no need for external cooling when the light doesn't heat up or give off much heat  

I got over 300w in a 1mx1m tent, bar style LED that covers the canopy with diodes from wall to wall, more than what's needed in that space, but in the right spec atleast cause I only flower in that tent so never need to accompany vegging plants under that light and I can toggle W output for early flower - late flower output increase, though I've had it on full blast during peak summer and we get 40°C days down here in the Winelands, those hot summer nights with my lights on I hit about 25 to 30°C inside the grow room and lights off during daytime keeping things dark I still sit at around 25 to 30°C inside the tent. pretty constant. my grow is in a insulated wendy house outside in my yard. never had any heat issues and I got no aircons no external cooling none of that. 

Actually thinking about running a small heater on a timer in there during the colder months, cause I am not a fan of purple weed. I don't smoke with my eyes.  

 

[koosjr]

Well, the big problem is I will not be able to change my light cycle to work at night when it is cool...

I simply will not have the power for that. My power is in the day when the sun is shining.

Regardless of that, I would like to see your light design that does not give off any heat.

LEDs, even the best of them is about 50% efficient - unless you can show me products where this is not the case. Shit, I have looked at many data sheets of components and there just ain't that magic one in there.

Sorry, it simply is impossible. A 1W LED will give off about 0.5W of heat minimum and if that LED is inside your room, it is room load.

I am not trying to be wiseguy here. This is just plain and simple realities of engineering. I have never come across a 100% efficient device.

As said, the fact that your hand does not feel the heat, does not mean there is not the same heat. It is just spread over a larger area.

I also see people are struggling to heat up boxes in winter - and yet a simple layer of good bubble wrap sheets around the box will probably more than half the heat one would need to keep it hot. In that case, the light alone will be enough to heat it up.