CXB3590 1500W

[Megalomando]

I guess I need to assemble and then give the panels a test run to see how much heat it generates in a still room. While I can put a fan at the top of the tent to make for better convection at the heatsinks, if the best thing is to add a fan, I can do that, I was just hoping to simplify things by eliminating the fan if possible. - One plus about having everything on one board is the ease in using it in more than one place, sometimes it's helpful for things to be self contained.

One possibility I see is much like the photo Supra offered of the 3.5x40 version with the driver on top & a fan present. If I were to get thin aluminum sheeting and use "L" tabs, I could anchor it to the top of the fins and extend it the full length of the heatsink. With a hole for the fan, a 120mm fan would pass air to all the fins and eject it out both ends. Like in that photo, the driver could live on top. My tallest point will be 78" so any space saving in the lighting will be a help. Thankfully I won't have to deal with the heat from a HPS. If the passive temps are not much with the 4 50W COBs, I might could use pressboard instead of that aluminum sheet, that will be easier for me to find than aluminum sheeting. If I had the right tools and found the aluminum, I could make a U-channel to fit on the top of the heatsink. But all I have are hand tools.

Years back I had something that would be perfect, it was a military surplus heat sink of long Around 30" long x 8"-9" wide extruded aluminum. There were two halves that made it as one. They slid together lengthwise, locking as the two edges linked together and when slid together, formed a hollow heatsink with the fins inside and a smooth rectangular surface outside. A 120MM fan fit perfectly in to one end and blew the heat out the other end. Banks of power transistors were on "shelves" on each side. I had two of those, let them go cause I just didn't need them then...

 

[SupraSPL]

So I was playing with the inverse square law formula and it doesn't really seem to be working for the distances we need. For examle, let's just forget that we are talking about leds and just think of it as numbers, if your input strenght S = 1600 and you enter 6" for "r", converting "r" to meters we multiply by 0.0254, so here what happens
I = S / (4 * pi * r^2)
I = 1600 / (4 * pi * (6 * 0.0254)^2) = 5517
I get a bigger number than the input. This happens when I enter small decimals for distance. It seems to work straightforward with distance being 1+ meters. I'm sure there's an explanation....

So the reason that is happening is because 1600 PPF emitted in a globe pattern does equal 5494 PPFD at a distance of 6" (.1524 meters), because the light would be illuminating an area less than 1 sq meter. So no worries, your formula is working correctly. If you doubled the distance to 12", that same light source would result in 1374 PPFD.

 

[mc130p]

So the reason that is happening is because 1600 PPF emitted in a globe pattern does equal 5494 PPFD at a distance of 6" (.1524 meters), because the light would be illuminating an area less than 1 sq meter. So no worries, your formula is working correctly. If you doubled the distance to 12", that same light source would result in 1374 PPFD.

Since the LED is directional, you might get better rough numbers by utilizing half the solid angle instead of the whole 4pi.....because essentially half the sphere is gone. Are you guys doing this, or using the convolution of the angular dispersion as well for PPFD? Mean value theorem says you can use the average and get PPF just fine. Sorry for the interruption...I haven't been following too closely.....

 

[SupraSPL]

The example above was for a theoretical light source emitting in a 360 degree globe. You are right the formula would have to take beam angle and beam characteristics into account for an LED.

What might be useful for the calculator, estimate average PPFD based on the COB setup and the canopy space it will be covering, then generate a distance to canopy and COB spacing suggestion for each design. We can now easily back that up with in-canopy PAR measurements to verify the recommendations are close.

 

[mc130p]

who's working on the calculator?

 

[alesh]

It seems to me that overlapping and reflective walls will make up for not having spots with less light. As of the intesity, it will be different at different levels. We can go back to the the average PPFD but it will be a different number.
By the way, can you post the trigonometry formulas, let's look at them?

Basic concept would be to find out area that a source covers at certain height (takes height and total included angle). Divide this area to as many smaller ones as possible and calculate averaged PPFD for the smaller areas (takes height, angle and the data from the radiation pattern chart). Sine & cosine law should be enough. Shouldn't be very complicated to implement.

 

[Greengenes707]

Basic inverse square law applies to a POINT SOURCE ONLY. Anything else will require much more math to accurately determine dispersion.
And like pointed out before in here...confining your grow room limits it beyond usability.

PPF of a lamp is what matters. Then using the correct amount of PPF per area(PPFD) is what will give you the desired/corresponding results.
Distance form the source is only making things more or less "comfortable" for the plants when referencing PAR meters. The lamp is still emitting the same amount of light(PPF).

Put a 1500w heater in a room...the 1500w is like PPF
Room=canopy, you=single plant
Stand next to the heater and it's "hot"...like being right under a big light with higher instintaneous PPFD readings.
Stand on the other side of the room and you barely feel the warmth like having low PPFD readings.
But the whole canopy/crop is still only getting 1500w(PPF) no matter how you disperse it.

If working with a fixed amount of light(all indoor is), it's up to the grower to make the most of it and disperse it as evenly as possible.

 

[bggrass]

So, I'm attaching a test version with an input for distance in inches. The calculator then takes the PPFD value from couple boxes above runs it together with the distance through the I = S / (4 * pi * r^2) formula and displays average PPFD "I" at that distance. Now this could be way off but it's a start. I can add an input field for desired PPFD and then the calcuator can take that value use the average PPFD from couple boxes above and display the distance necessary. This is easily done. For example if we use the I = S / (4 * pi * r^2) formula and considering we have S from the cob configuration and we also input "I" as a desired PPFD and then we solve for "r". r = sqrt(S / (4 * pi * I)). The hard part is figuring out what formula to use. Now considering Greengenes point source only comment and alesh's trig suggestions, I thought that since we know that cxb3590 surface is 30mm and (I think) it beams at 120 degrees we take the calculated distance that we add to the distance from canopy to adjust for point source. Here's what I mean tan(theta) = opposite / adjacent; adjacent = opposite / tan(theta); adjacent = the distance to add to adjust for point source. I'm attaching some of my scribbles, hope they make sence.

Another thought is, since we can measure PAR to verify if the formula is correct, why don't we adjust a formula for PAR over distance and then we can just multiply by umol/s to get the rest of the numbers. As of cob spacing, that can be done as well, as long as we come up with a formula to it. I think we should take it one at a time though cause it might get too confusing.

As far as trig functions, they are pre-set in the language and can be easily used if incorporated in a formula.

https://www.dropbox.com/s/fvxlyq31b5r6btt/testCalculator.jar?dl=0

 

[bggrass]

Another thought is, someone with a par meter can just take the readings for a single cob at 1 inch, 2 inches, 3 inches........ 30 inches and we can just plug a graph with distance and par values and that should give us the function for par over distance. If it gets done at 90* for straight under the cob reading and then let's say at 60*, then let's say 2 cobs par values at 60* give us the same number as 1 cob at 90*, that will pretty much give us the formula for distance between cobs for even light over canopy.

I'm thinking this all theoretically.

 

[SupraSPL]

I thought that since we know that cxb3590 surface is 30mm and (I think) it beams at 120 degrees we take the calculated distance that we add to the distance from canopy to adjust for point source.

Another thought is, since we can measure PAR to verify if the formula is correct, why don't we adjust a formula for PAR over distance and then we can just multiply by umol/s to get the rest of the numbers. As of cob spacing, that can be done as well, as long as we come up with a formula to it. I think we should take it one at a time though cause it might get too confusing.

The beam pattern of COB is not straight forward 120 degrees, it emits in a cone that gradually drops off, all the way up to 180 degrees. That pattern would have to be taken in to account to estimate the effect distance would have on the PPFD.

It probably would be easier to work backward as you suggested, take PAR readings with known reflectors/lenses and various distances. But it is even easier to aplly the 10% lens/reflector losses and divide the resulting PPF by the desired canopy size. Distance comes later in that process because it depends on how you choose to space your COBs and what reflector/lens you choose. A common misconception, distance to canopy is not the design variable used to adjust PPFD.

For example, if you hung a lamp at the "sweet spot" distance it was designed for, and then you increase the distance to decrease PPFD, a lot of the light would end up on the walls. If you decreased the distance to increase PPFD, you would have a bunch of hot spots and dim spots, poor uniformity. But if you left the lamp exactly where it was, you could reduce or boost the PPFD using the dimmer, with no height change necessary and maintain the same uniformity.

 

[bggrass]

I think if we have the par numbers of a cob at miltiple distances and for every distance multiple angles, that will give us enough to estimate different configurations. Don't see a reason to deal with inverse square law after that as we have the actual numbers. I am starting to understand what you mean by once the system is designed all you can do is dim it. Otherwise it all gets thrown off.

 

[bggrass]

There might be no need to take readings for every inch and every angle. All we might need is 3-4 readings at different distance and 3-4 different angles. When plugged in a draph, hopefully we'll get a straight line and that can give us the numbers at any distance we want and any angle.... if we get the straight line...

If it doesn't come as a straight line it's still posible, just might take little more readings to estimate the curve.

 

[SupraSPL]

Most recent batch gave me the best numbers I have been able to manage so far, 1.47 gpw in 56 days. COBs were running at 20-25W ea, 55-65% efficiency. A mixture of CXB3070 3K AD, CXB3590 3K CB and CXB3590 3500K CD. Max intensity was ~ 700 PPFD averaged

I put DIY reflectors on the COBs that are on the edge of the canopy. ROLS soil mix that was mostly no -till. I did use a lot of EWC during transplant and I watered in a bubbled few nutrient teas during flowering. Canopy temps were 75-78F. CO2 ranged from 400-900 PPM.

I was amazed how fast they finished and I assume that is due to the lower canopy temps now that summer has passed. I figure 60 days is an average mark for how fast most hybrids can finish. So if it finishes faster, I consider it a bonus to gpw and if it finishes slower I consider that a penalty to gpw because I want to move the next batch of ladies into that flowering space as soon as possible.

So if the numbers are adjusted with the 60 day mark in mind, some of my previous best looks a lot lower because of higher temps and makes the 56 day finish looks even better:
1.34 gpw in 70 days = 1.15 gpw in 60 days
1.47 gpw in 56 days = 1.58 gpw in 60 days

Huge difference in productivity when I look at it that way. Also worth mentioning the quality of the 56 days batch was much higher than the 70day batch, also due to lower canopy temperature I assume.

 

[ttystikk]

Most recent batch gave me the best numbers I have been able to manage so far, 1.47 gpw in 56 days. COBs were running at 20-25W ea, 55-65% efficiency. A mixture of CXB3070 3K AD, CXB3590 3K CB and CXB3590 3500K CD. Max intensity was ~ 700 PPFD averaged

I put DIY reflectors on the COBs that are on the edge of the canopy. ROLS soil mix that was mostly no -till. I did use a lot of EWC during transplant and I watered in a few nutrient teas during flowering. Canopy temps were 75-78F. CO2 ranged from 400-900 PPM.

I was amazed how fast they finished and I assume that is due to the lower canopy temps now that summer has passed. I figure 60 days is an average mark for how fast most hybrids can finish. So if it finishes faster, I consider it a bonus to gpw and if it finishes slower I consider that a penalty to gpw because I want to move the next batch of ladies into that flowering space as soon as possible.

So if the numbers are adjusted with the 60 day mark in mind, some of my previous best looks a lot lower because of higher temps and makes the 56 day finish looks even better:
1.34 gpw in 70 days = 1.15 gpw in 60 days
1.47 gpw in 56 days = 1.58 gpw in 60 days

Huge difference in productivity when I look at it that way. Also worth mentioning the quality of the 56 days batch was much higher than the 70day batch, also due to lower canopy temperature I assume.

Soooooo modern COBs, mid fifties efficiency, 1.5gpw and in organic no till soil.

I'm impressed.

 

[threetwobravo]

Most recent batch gave me the best numbers I have been able to manage so far, 1.47 gpw in 56 days. COBs were running at 20-25W ea, 55-65% efficiency. A mixture of CXB3070 3K AD, CXB3590 3K CB and CXB3590 3500K CD. Max intensity was ~ 700 PPFD averaged

So can I take away from this a average of 700 PPFD played a good part in getting 1.47 gpw? Could this average of 700 be applied to any size grow?

 

[ttystikk]

So can I take away from this a average of 700 PPFD played a good part in getting 1.47 gpw? Could this average of 700 be applied to any size grow?

Yep. The game then becomes about how large you can have your canopy and still be that efficient.

 

[SupraSPL]

Yes I thought 700 seemed like a good compromise in terms of intensity and efficiency, although if your ladies are tall with a 3D canopy you would need more PPFD. That 1.47gpw turned out 1.81 units from a 4.5'X5' canopy, ladies were 3.5-5.5ft tall. You could scale that up as large as you need. Even better if you could use a shelf system and double up the use of floor space, which I think you could do with as little as 10' ceiling height.

The canopy would have intensity that ranges in this general area.

 

[SupraSPL]

The Grease Monkey ladies are about 6 weeks.

This one has lots of frosting coming on but seems to have very low aroma and yield so far


This one seems to have decent yield potential but so far also low aroma


This one has very strong sweet smell, really nice, plenty of greasy frosting. It does has some density coming on but still seems like it will be a low yielder from what I can tell so far.

 

[SupraSPL]

For comparison, here are some decent yielding ladies in the same space @ ~6weeks All super stinky

TH Seeds - Critical HOG


TGA - Ace of Spades


G13 Labs - Sour Jack

 

[yoda954]

Hi

I like to know if it is possible to put a interuptor instead of potentiometer 100 ohm driver "HLG-185H-C Series 1400" to put my panel vegetative way ???
example:
interuptor on mode 1 = max power of 1400 amperes panel
interuptor on Mode 2 = medium power 700 amps or less
how can do to put a interuptor to do this or should I get another driver?

thank you a lot
have a good day