ChiefRunningPhist
Well-Known Member
Ya I probably, didn't need to create an entire thread, I should have just looked at the ISL a bit more to determine UVB chip needs. Lol 
(μmol/s)/(PPFD) with respect to Hang height
- Thread starter ChiefRunningPhist
- Start date
-
- Tags
- hang height ppfd umols μmol/s
Moflow
Well-Known Member
ChiefRunningPhist
Well-Known Member
Lol u da man, im just sitting here like
Thundercat
Well-Known Member
Excellently summed up!
wietefras
Well-Known Member
OMG you seriously are putting any credit to this nonsense? Come on man, you should know better! FFS! I'm truly disgusted.
Especially the notion of deeper penetration due to higher hanging height is sickeningly stupid. Seriously! The reason that idiotic calculator shows better penetration is because you would need to start with 64 times as much light for the 40 height as you did with the 5 height (8 times the distance means 1/64 intensity). So according to inverse square law he needs to start with 32000umol to get 500umol at 40 height vs 500 umol at 5 height. Yes of course 32000umol will give you great "penetration"! He's basically comparing a candle to a stadium flood light and then concludes that the flood light gives better "penetration". No shit sherlock!
Either way, it's not that inverse square law does not apply as a law (per individual light point), but the point is that due to overlap and reflection you do not see it when you hang the light at the correct height. Also light does not disappear. The total amount of light stays the same (unless it's absorbed!), it just gets spread out more and of course the local intensity will be less.
All that means that inverse square law does NOT apply for the height of a whole fixture in a grow tent. Inverse square law only works when the light can spread over a quadratic larger area. This is impossible in a reflective tent, because the light reflects back. It cannot spread further than the walls of the tent. So inverse square CANNOT apply once the light starts bouncing back.
The problem with these matrices is that they are not based in reality. No one would hang a single 320W QB in the middle of a 5x5 tent. So indeed the light gets to spread much further than it normally would.
Still even then it's 100% clear inverse square does NOT apply. Take the green chart with 3 heights per spot. One 12" center measurement is 1310. At 18" the distance is 1.5 times as large, according to inverse square the light intensity should be divided by 1.5*1.5= 2.25 and be 582. It's measured at 830. Way off. At 24" (ie double the distance) it should be a quarter of the intensity as the light would have got to spread 4 times the area. Which would be 327, but in fact measured is 570. Again way of.
Just think about it. If at 12" you measure 1000umol do you really think you get only 62.5umol at 48"? Is the floor of your tent truly almost dark?
The reason light diminishes with increasing height in a grow tent is ultimately reflection losses on the walls. When you are close to the light (ie at less than the normal hanging height) you will see some exponential intensity drop, but once the light is distributed across the whole tent (ie around the actual height you would use), you will only see intensity reduction due to reflective losses.
ChiefRunningPhist
Well-Known Member
Seems your contradicting yourself here, using ISL to bolster a candle to a spotlight comparison, but then arguing against it later.
You've made some good points about reflection affecting beam angle though and the phenomenon of a tent acting like a giant reflector hood maintaining light within.
If you're measuring a PS at 1000μmol/s and then 48in away you're measuring 62.5μmol/s, then the 1000μmol/s is just distributed over X amount of greater area, and that area increase is in a proportion relating to the distance measured from the origin. So if you're at 62.5μmol/s at 48in away from a 1000μmol/s PS, then you're emitting the PS 1000μmol/s output in a spherical pattern. An LED does NOT emit in a spherical pattern.
The power doesn't change, the number of photons stays the same, only the area that the photons occupy changes. So its a matter of how fast the outer surface area is expanding and the initial intensity, compared to the amount of distance traveled from the origin.
This part seems pretty straight forward, although I'm still trying to comprehend the unit multiplier from the explanation posted.
My thought is that you can model LEDs after lasers, just disregard the Raleigh range, and make sure to enter the units of θ as radians, and as the half angle. Though I still think I'm missing something.
Also, if the emission projects in a cone, but cuts through the canopy in a planar way, the center of emmision will have the least distance traveled whereas the outer edge of the planar emission will be the greatest distance from origin. I'm thinking that the edges will change rate of intensity differently than the center will change rates of intensity when measuring from a plane which is tangent to the emmision sphere, or when measuring PPFD at canopy. What I'm saying, is If you wanted to model the ISL with a par map, that you'd have to shape the par map to the emmision shape, ie so all points were equidistant from the origin, so pull it up around edges and more bowl like.
What you said about reflection affecting intensity or ISL is pretty significant I think. It seems that the tent walls essentially stop the beam angle from expanding, so youd have to add what would have expanded back onto itself which is essentially decreasing surface area, or breaking the consistent expansion rate. Good point.
I'm only trying to determine UVB chip needs and they will be on a dimmer, so as long as I have an idea of what to expect in intensity drop over distance, then the added reflection can be mitigated by dimming.
Randomblame
Well-Known Member
Bro, you know I know all that shit. I was part of the discussion. I've only meant it as example and have also mentioned he was an idiot. Its pretty obvious that you need more watts to reach same intensity level at canopy level with more distance and that you get higher numbers down below using more watts.
If you remember he has created a few other tools too. One of them was just calculating par numbers at different heights and that tool has worked and the calculated results were close the the sensor readings.
To stay honest, it was not all complete nonsense what he said; the basics were there. He just drew the wrong conclusions and misinterpreted things permanently. (See tool above)
Probably because he was all time drunken and has rather preferred to offend and insult instead to learn something.
Chief is just lookin for an easy way to calculate intensities with multiple LED's at different heights using the ISL. I've no idea if the tool is still online but to just calculate par numbers of strips at different height it has worked.
@ChiefRunningPhist
Here is another tool you could use but you can only add total output, the length of the fixture, round or rectangle grow area and a few other things and its not only for LED's. But its based on ISL and gives pretty accurate results. But I don't know if you can transfer it to uv light somehow.
http://dev.edman007.com/~edman007/pub/par-dli-cal.html
Last edited:
Randomblame
Well-Known Member
He had a few tools and one was just calculating PAR numbers of multiple LED strips at different hanging heigths. If its still online its maybe helpful because it factors in intensity at every 1° beam angle.
But be careful... He often draws wrong conclusions and misinterpret things.
I believe these one tool was the only one that has worked.
ChiefRunningPhist
Well-Known Member
Well I don't know the guy but I'm guessing he didn't give a fuck lol anyways thanks @Randomblame I'll take a peek
Grow for fun only
Well-Known Member
Hey professor,i see your thread and got the useful info and i do have a question for you. In the earlier day one guy sent me the info about his led, i see its PPE such high which i cannot believe it, since so far I only see 2.35-2.79umol/j for some led in the market,so i would ask if you guys ever have chance to know it or can give me some information. thx alot.
ChiefRunningPhist
Well-Known Member
Depends on the chip, drive current, and operating temp, but you're probably at 2.15μmol/J if it's LM301B at 800W and what I'm assuming is a 4x4, unless it's high effeciency binning LM301B, then you might be 2.4μmol/J at 800W.
The SPD doesn't look like its added any red which could further increase effeciency, though again depending on chip selection drive current and thermal mgmt.
The new hlg is blasting ~2.6μmol/J with added high effeciency reds to bump the total effeciency even higher than if they used only the high effeciency bin LM301B they run. So a knockoff is most likely not using high effeciency bin LM301B, and it's not using red which could have had the possibility of boosting its effeciency, but even then if it did include red, they'd most likely be less effecient than the white they were running which would pull the 2.2μmol/J down to maybe 2.1μmol/J.
Those are just guesstimates but you can for sure know that the 2.9μmol/J is bullshit. Probably 2.3μmol/J at best is my guess.
Its a nice distribution, and it will grow the plant well, it's just not hitting 2.9μmol/J.
ChiefRunningPhist
Well-Known Member
Also PPFD is not soley dependent on PPE. Furthermore, determining real world PPFD diminishing rates with respect to height, initial intensity, beam angle, and enclosure, is a lot easier utilizing a computer that can handle thousands of data points. I think this is what IES files are and what TEK was talking bout, using a computer to model the real world incident intensity of whatever fixture you're trying to analyze. I've not heard many growers talk or use, so not really sure the implications, but if that's not what they are then I think entering all these data points that brother @Moflow supplied into an excel spreadsheet, and then average all the fixtures intensity drop off rates per beam angle per height per initial intensity per initial area of illumination, you can model X design for Y intensity @ Z height
Last edited:
Randomblame
Well-Known Member
It's possible to create something that reach 2,9μMol/J but you would need a hell lot of diodes which makes the fixture much more expensive. Quantumboard v2 diodes run at ~175mA or so and reach 2,5 without deepreds. To get 2,8 you need to run them close to test current which is 65mA, maybe 80mA is enough. And you have to use the exact same top bins. Because HLG gets selected top bins its not easy to find them if not impossible.
The SPD curve looks pretty similar to the QB redspec and they have not added much deep-red. So in the best case they increase efficiency by the same 0,1 which would mean 2,9 total.
Thing is, can you trust the seller/manufacturer?
Do they really use more expensive top bins?
Which deep-red diodes do they use? SMD5630's in deep-red would not improve efficiency. It needs at least top bin Prolight or Samsung deep-reds, Cree XPE2 or Osram Square would be even better.
They should at least have an independent lab test and not just a par map and a few numbers. If its something you can find on alibaba I would not trust them. Even for a DIY'er such a light would cost +1000$ build from quality parts.
ChiefRunningPhist
Well-Known Member
Actually, you only need 1 diode to hit 2.9 μmol/J.
Lol, In the case of his scenario though, 800W in a 4x4 at 2.9 would be the ceiling covered wall to wall at 0.5in2 spacing per chip 4,475 chips total, or 16 QB280s per 4x4.
Lol, In the case of his scenario though, 800W in a 4x4 at 2.9 would be the ceiling covered wall to wall at 0.5in2 spacing per chip 4,475 chips total, or 16 QB280s per 4x4.
wietefras
Well-Known Member
No, it was all complete bullshit. Just stop. ISL does NOT apply in grow rooms for entire fixtures and therefore his tools and opinions were all useless.
Also, you should not be adjusting PPFD by hanging height and all these matrices are really useless.
The correct hanging height is solely determined by light distribution uniformity. Nothing else. You need to spread the light to the right amount of uniformity over the whole area. Professional growers use the rule of thumb that the darker regions should still get 80% of the average light intensity to prevent uneven plant development.
The optimal hanging height determined that way is a constant for a certain setup. If you then want more or less light you need to adjust the power.
If you lower the height to increase light intensity, you will get a hotspot and dark regions (and uneven plant development). If you hang the light higher to reduce intensity you are simply wasting the light on wall losses and you should be dimming instead.
Once you have the light at the optimal height, going a few inches up gives you a inverse linear relation of intensity to the height. Depending on the size of the room you lose for instance 1% or 2% of the light per inch of extra height.
wietefras
Well-Known Member
No, you misunderstood. I'm showing how his example would work out if ISL actually applied like he claims it does. Or in effect how idiotic it is.
Either way, That nofsgiven chart that Randoblame posted is complete bullshit. Again, I'm truly flabbergasted that he posted it at all.
So? The circle projected from a led grows quadratic with distance.
What it boils down to is that the light expands a sort of ISL like relation until it starts hitting the walls. That's the point where you start reaching uniform lighting distribution over the whole surface. From then on you will only see a pretty much linear decrease in intensity caused by wall losses.
You really shouldn't think of it as intensity drop over distance. It's light spread over distance. You still have the photons (until they are absorbed somewhere).
You need to calculate how many photons you need. Take the desired UV intensity and multiply by the surface area. With correction for wall losses (between 10% and 40% depending on how much of the wall is lit up). Then determine how to distribute that over the whole surface. That will be the hard part. It's easier if you have a bigger beam angle. For instance a 150 degree beam angle spreads the light over 4 times the area compared to a 115 angle led. That's how I combined 4 Osram FR leds with 9 COBs. 4 light points would normally need to hang higher than 9 (for uniformity), but due to the bigger beam angle of the 4 they can then hang at the same distance for similar uniformity.
Beam angle and also light source count is what matters for uniformity. Plus the height of course.
For COBs and single light points you can go with 2/3 of the maximum distance between them. For example if you have 4 light points (with 115 degree beam) angle in a 4x4 then the light points would be 24" apart and the needed hanging height would be 24" * 2 /3 = 16". You really don't need to calculate ISL or anything. It simply just scales up with the size of the tent since the beam angle is the same.
Last edited:
Randomblame
Well-Known Member
Again, you know I know all this as we have discussed it often enough!
Inside a tent you lose some canopy intensity with more distance but you don't really lose much light because up to 90% gets reflected and only a little gets converted into heat or is absorbed.
But its still a calculation and you can also factor in reflective walls if you want.
Dialux for instance is a software that can also factor in reflective walls and its based on ISL.
Even though NFTG has misinterpreted a lot (if not all) this one tool has worked pretty well! (even a blind chicken can find a grain every now and then)
What Moflow and I've posted is not the tool I'm talking about. I only used it to show that it was actually a tool and not just a picture/screenshot. That this tool is just useless and an example that shows his lack in understanding is another thing.
But even if he was wrong all the time his ISL calculator has worked! At least the numbers were pretty close to the sensor readings.
Thing is, Chief is actually looking for a way to calculated how much UV diodes he needs to get a certain intensity at a certain distance. He just want's an easy formula or calculator to get some numbers to work with.
Reflective walls doesn't matter much with UVB light cause most plastics, paints, clothes, glass and also plastic coated mylar would not reflect it. That means he can simply use the ISL to calculate the intensity at different heights. No need to factor in reflective walls. The amount of reflected UVB is probably lower than 10-20% inside a tent.
For that reason I've thought he could use a simple tool that calculates intensity at different heights without reflective walls.
And as I've seen that screenshot in one of Moflow's postings I remembered he (NFTG) has created a few other tools as well. That the ISL tool was the only tool that has spit out correct numbers is written on another paper.
It was really not my intention to give NFTG some credits for his work or better his wasted time.
I do not even know if it's still online...
But with such an easy calculator tool he could just enter μW/cm values instead of lux, lumen or μMol/s because also UVB distribution is based on the same ISL.
Hope it's clear now.. Shitty language barrier..
wietefras
Well-Known Member
Yes, I have written a simulator that does the same yes. It does millions of "ISL" calculations. However, the point is that for the height and distribution of light of a FIXTURE (as in an "array of distributed light points") as a whole, Inverse Square law is simply irrelevant.
Mylar and plastic doesn't reflect visible light much either. So what? It's the aluminium coating which does the reflecting and that's still well over 90% for UV. You lose some on the carrier material, but still.
You mean too suggest that he's going to manually do a simulation of a light distribution of multiple light points over a surface? Sorry man, but come on.
Anyway, it's not just reflectivity that is at work. In general it's overlap. That might come from the reflection overlapping back, but also overlap from the distribution of light points and the uniformity that creates.
For instance, 4 light points project 4 cones and these overlap. This overlap already cancels out the effect of inverse square law, because as the light bleeds away from one light point the light from the other three light points overlaps back.
The rule of thumb that I worked out for COBs (and another one for led strips) is just as valid for UV leds. It helps determine the hanging height and from there on the only difference will be in wall losses. If you have no walls those will be bigger, but that's pretty much irrelevant.
A good hanging height is 2/3 of the (maximum) distance between the (115 degree beam angle) light points. Or if there is only one light point then it would be 2/3 of the width or depth of the surface area (ie in e 2'x2' tent it would be 24*2/3=16") .
At 2/3 of the distance between the light points, only the corners are below 80% of average intensity, but then the corners are almost always a little below.
This is a simulation of a 2x2 matrix (4 light points total) at 2/3 of the distance (blue is below 80% of average):
It actually doesn't matter how big this simulated room is. It could be 2x2 or 4x4 or whatever square dimensions. The height requirement scales with the size of the room and then the overall light distribution stays the same.
You really only need to think of how many umols/s/m2 of photons you want and then multiply that by surface area and add some for wall losses.
Inverse square calculations do not come into play. Nobody does that when they buy a led light. Do you buy a led light for your grow room and then start calculating inverse square per light point? No of course you don't. You have a 1m2 room and you want a PPFD of 800umol/s/m2 average. So you hang say a 900umol/s PPF light in there and when it's on the correct height and the light is distributed uniformly you're done. Nowhere have you used ISL in that.