yes, in this case I do. We always ask the grower how much light he requires when we make a lighting plan for a customer. Most growers however don’t have a clue and just ask for 1000 umol s-1 m-2, which, in the right climate room with the correct cultivars is fine, but not with every cultivar or cultivation method.
I just don’t understand that they keep doing that. I would recommend growers with little experience or no CO2 no higher than 600-700 umol s-1 m-2 and that might already be too much for them.
And then you see those growers who hang their DE fixtures in a “5x5” (or worse, 4x4!) and find strange that they have these huge light peaks and sub-optimal plants.
This is one of the most asked questions at trade shows: “So my buddy tells me that they should be hung in a 5x5/4x4 grid, right?”. Sorry mate, your buddy don’t know s***
That has been a big question since I’ve started looking at lighting for cannabis. If I interpret a plant’s response to incident light as a resonator, at some point I would over drive the system and plants would stop taking in light for use in photosythesis, similar to a camera becoming over saturated and showing white in images.
Is there any data on a “rule of thumb” for minimum/maximum PPFD values that cannabis plants need? (obviously strain by strain would change these values) I’ve seen a few studies on this for tomatoes and lettuce. Otherwise it feels like it’s a bit wild west with the light output from lots of these systems. I’m trying to understand the community metrics and standards as best I can.
Yep, there is a limit to the amount of photosynthesis a plant can do, and it is determined by species, as you said.
The difficulty is that every species (and cultivar) has a unique limit of where their maximum rate of photosynthesis is based on their evolution and genetics. And I haven’t seen any body of research on specifically cannabis’ photosynthetic limits. It’s difficult to measure metabolism, and so most researchers use plant biomass as an indirect measurement.
The closest article of interest I can find is from this paper:
The problem with this number is that it’s a maximum (before output drops) and doesn’t ask how efficiently that temperature and light output can be reached. Having a lower PPFD might be more efficient, cost-wise. But that’s a business question barging into pure SCIENCE!
There are a few more researches as to the saturation point of Cannabis. Generally they range from 1400-1600, but as you know specific cultivars take more or less light, and it doesn’t take optimal DLI into account.
In general though you will see a decrease in added photosynthetic efficiency at higher light levels as plants become CO2 limited. Adding CO2 will bring the curve a bit up, and lower temperatures allow for a higher ppfd (at a lower efficiency!).
In general though you can say that as from 600-700 umol s-1 m-2 an increase in light level will not give you a linear extra photosynthetic response. Now for a low value crop this will have more impact on your decision to increase the light levels than for a high value crop, where you also need to take into consideration that, unlike many plants such as tomatoes and roses, cannabis is a short day plant so all light needs to be within a 12 hours period.
It’s a pity we can not have a good discussion with some of the other manufacturers here other than a hit and run type of response, possibly just aimed at keeping me of my work ;).
I like a good discussion and I am not an LED hater at all. I am just trying for manufacturers to specify their eiquipment correctly and not making obvious wrong claims, as it hurts the lighting industry as such and sends out wrong signals to aspiring buyers. In the end, it is in nobodies interest to hype any technology.
Just jumping in now, but curious how you all appear to be having a great technical discussion but you state that labeling a 380 nm peak UV led as not UV? Based on everything I have read anything below 400 is UV. Of course there are classifications of UVA, UVB, and UVC but anything below 400 nm is UV so saying it is hardly UV seems a bit misleading. Can you elaborate? And our multiple different peak wavelength LEDs are glass ones, which as you point out are amazing LEDs and their performance has come a long way since we started using UV 7 years ago.
Also, since you seem to be one of the more knowledgeable people, especially with regard to HPS, can you tell me why all HPS companies use integration sphere numbers with a bare bulb instead of measuring the total fixture output with the bulb in a fixture. At least all the readings I have seen. I don’t hate on HPS, hell I grew with them for years before LED and had a great time with it, just curios why they don’t use real-world numbers based on an actual fixture. Especially when it is not hard to throw the whole fixture in the integration sphere and get an accurate number, assuming one has a large enough sphere.
This is such a great discussion and topic!! This is so much of what I am researching daily. Regardless if you everyone agrees, I think the collective wealth of knowledge on the topic is fantastic for the industry. I am truly enjoying reading this thread and being able to research both sides accordingly.
Yes. The effect of 380 nm compared to blue light is insignificant. 380 nm have been around for a long while, specifically in aquatics fixtures, for visual popping of the colors of the corals. However, they do not have much added value for medical plants. Many manufacturers put in on their data sheet to have it in, but really, do the tests and see what 380 nm does to your plants.
Where it really becomes interesting is when you more more to UVB. That’s a completely different beast as in effect on plants.
I was just in the studio Friday to record an interview with a Philips lead engineer from their lighting lab, the largest in Europe. I will make a long story short, this is what it comes down to:
The Lumens output of a HPS varies with the temperature of the lamp, this is specifically an issue for the DE lamp as it is nitrogen filled
So a naked burning lamp can read more lumens than a lamp burning in a reflector
However, though the lumens value can differ 5-10% (!) the umol output does NOT (less than 1%)
All ANSI labs measure lumens from fixtures to calculate efficiency, which ,as you can understand now, will give better readings for open reflectors for example. However, in reality there is no, or no significant difference anyways, between the output of a naked burning lamp and and a lamp in the fixture.
To calculate the output of a HPS fixture you must take into account the reflector efficiency which is measured based on the a sphere measurement and a photogoniometer measurement (for which we use, together with Philips, use a special lamp and protocol). So actually those comparisons which you see published by some of our competitors don’t mean a lot.
To calculate the efficiency of a fixture you need to take reflector efficiency and ballast overhead into account.
It goes even a bit further than that if you really want to calculate with the light values: You need to take into account the light maintenance of the lamp over a year to calculate the average output. The same goes for LEDs btw over their lifetime.
An DE HPS lamp (at least when you have a good one) depreciates about 3% per 5000 hours max. A bad lamp does 15%. So with a good lamp you have average 1-1.5% average less light than calculated from a new lamp.
If you have an L90 (light maintenance in % at rated life) with LEDs at 30.000 hours, at 15000 hours you probably have 5% losses, which would also be the average losses over lifetime. Now a HPS light you can replace easily, but not so much the LEDs in a fixture. So when making a horticultural lighting plan you need to take the projected use into account as well.
In our light calculations, which we provide for every commercial grow and committed customer for free, this is already taken into account.
As I am traveling to the US for 2 weeks the video will be out in about 3 weeks, till then already take a look at part 1 and 2 of the series:
@Theo: Could you clarify something for me? I understand that as the temperature of the lamp changes, the luminous flux changes since temperature shifts the blackbody spectrum accordingly. Thus, “perceived” light output or the lumens change. However, I’m unclear on the reasoning that a naked burning lamp reads out more lumens than with a reflector?
In an integrating sphere, the ideal case is that any source placed within the sphere will have 100% of the output light recorded. Obviously this isn’t the case but not terribly far off. So how does a reflector reduce the amount of light/lumens seen by the integrating sphere? Thanks!
Hi, Brian here from Transcend Lighting. Like some others in this chain, we design and manufacture lighting. Our focus is on LED lighting although our engineering team has a lot of experience with other lighting technologies used for both growing and general lighting.
To help clarify some of the testing procedures, it’s worth mentioning that LED lighting systems are measured using Absolute Photometry and other lighting systems, like HPS, tend to be measured using Relative Photometry. This is generally dictated by IESNA (Illuminating Engineering Society of North America).
Absolute Photometry is when you are measuring a complete system including electronics, light sources, reflectors and lenses all at the same time. This gives the actual performance of a system at a given set of ambient operating conditions.
Relative Photometry means that you measure various elements of a system separately - like a lamp and a reflector. This works well for systems that are well defined like linear fluorescent or HID where
total output = (rated lamp output) x (ballast factors) x (luminiere efficacy)
You can use any compatible lamp, with any compatible ballast, and any compatible reflector and relatively accurately calculate the system performance. This doesn’t work well with LEDs since they are more complicated semiconductor systems where the electronics and optics and heat sinks are generally all integrated together.
Of course, you can always measure an HPS or linear fluorescent system with Absolute Photometry, and this testing is usually performed when you need to take into account variables like temperature or humidity in your application.
Regarding light loss from a reflector, with any lighting system, including LED or HPS, every time light comes into contact with a surface there is some light loss. If a reflector is 94% reflective, every time light hits the reflector 6% is lost. If light hits the reflector twice, then 12% is lost. Reflectors, even though they reduce light output are still important because they can direct light to exactly where you need it. In the case of HPS, this is very important b/c the light source is 360d and usually you want all of the light to go in the same direction. We use reflectors in some of our LED design when they are mounted many feet above the plants (think greenhouses).
My apologies. My reference to “obviously this isn’t the case” is that the integrating sphere is collecting 100% of the light emitted from the source. There are losses at the entrance port and other areas, which is what I was referring to.
@TranscendLighting : Brian, fantastic response and great to see another person from the school of Optics in the space. I grew up in Rochester and know UR well! The reason I was posing this question is because I was wondering if there was a disparity from testing with a bare bulb or testing with a reflector. I know that the reflectivity coefficient determines how much incident light is reflected per bounce but does that account for all of the “loss” seen in the measurement using the integrating sphere? We tested a heated coil for use in some IR polymers in our integrating sphere and I had two runs, one bare and one with a back reflector but did not notice a 5-10% drop-off in output. Was really just trying to clarify where that loss was coming from.
@shibbyhockey04 Eddie, we actually have two Optical Engineers on staff from U of R. Lot’s of optical engineering with grow lights!
My guess, since you’re talking lumens, would be spectral shift of the lamp due to temperature. I didn’t think that HPS had much in the way of temperature dependence on spectrum, but a shift of a few nanometers at the edges of the eye’s luminosity curve can easily add up to 5-10% of lumen output. For umoles, the impact would be much less b/c all photons are weighting equally.
That’s what I was assuming was going on in the 5-10% loss. It’s been interesting to go from photometric units to radiometric units where I’m comfortable. Even looking at measures such as PPF where measurements are done based off photon count has been new and interesting.
I was reading through your technology and it’s really interesting stuff! The use of blue LED’s and phosphor coating to downshift into custom spectra is fantastic. Looks like you’re also getting a nice diffusion effect from particle scatter. When investigating the use of White LED’s,I read a few OSA papers on the degradation of the phosphor coating and what they were working on to improve it’s lifetime.
Could you possibly comment on what you’ve seen or experienced with the lifetime of the coatings and maybe talk about some challenges with phosphor coating in general? Really interesting application!
Unlike a sphere for small LED devices we use a 2m full sphere with the lamp centered in it. So the only losses are the holder of the lamp, which are compensated already when you calibrate the sphere.
A lamp typically gets a bit warmer in our HR96 reflector design, but specifically when compared to the more open reflectors. This doesn’t add any heat gain, but just a little shift of spectrum. Measuring in Lumens will give you a difference in output.
