Umol/Watt - "Impossible" Readings

kotobide said:

The lightscout meter is garbage.

https://www.rollitup.org/t/lightscout-quantum-par-meters.948811/

Click to expand...

I got one to test from the supplier at a trade show who wanted to know how their algorithms compared to Li-Cor. I then used their meters to do side-by-side comparison readings in sunlight, HPS and LED environments. I found that at lower umol readings (300 and below) the two sensors were nearly identical. From 300-600 umol the Light Scout showed about 3-5% less, and when you got up to the 1500umol+ range the light scout was reading about 15-17% less. So for me I don't mind using it for readings below 300 umol, as I know from my own testing it's nearly identical to the readings I get with my Li-Cor.

OneHitDone said:

Click to expand...

So my issue with this document is what's circled in yellow. They used a generic lumens to umol multiplier to generate PAR data. If you could use lumens to determine PPFD, our test light with Luxeon Rebel LEDs would have destroyed everything else on the planet... Yet it had 2.5x less PAR per watt than our standard units with far less Lumen output.

UL is insanely expensive (they want $17k to do the initial certification and I believe about half that per year thereafter to keep it active).

Does anyone here know of a testing lab (only one I'm familiar with is Light Testing Laboratories in Cali) that does this umol/joule test using an actual quantum light meter instead of a lumen multiplier?

HydroGrowLED said:

No point-source light distributes light evenly over an area. Light will always have the highest reading at direct center beneath a point-source fixture. Intensity will always lessen as you move away from center. Since we are talking about point-source lights here, even distribution is not a question of If, it's simply not possible. The edges of any point-source fixture will never match the intensity directly beneath the LEDs.

Click to expand...

That's correct if shining onto a planar surface from a distance. That's not what's happening in a sphere though. The sphere acts like a big optic or lens. If its light proof then the light stays within the sphere bouncing all around. Because its a sphere it bounces the light equally all over, or as close to 100% as possible. Then when you take a point reading you can translate that over the entire sphere surface area to get a total μmol reading.

HydroGrowLED said:

Light proof or not, zero is zero. The video shows 0 umol before the LED is turned on.

Click to expand...

2 things. 1 is that if the sensor isn't picking up the ambient, how accurate is the sensor? 2, if you're losing light while you're testing, ie you can see light, your results will be less than what they should be.

HydroGrowLED said:

This integrating sphere (8" diameter, 0.13m2) was created specifically for building precise, quantum-balanced spectral ratios to know relative umol output from one wavelength to the next.

Click to expand...

The ability to detect μmol of the individual WVs will be based on the spectral response curve of the sensor you're using, which I haven't been able to track down yet. So some WV may be picked up easier or stronger then others, which would screw with the results too.

HydroGrowLED said:

Just to be certain I'm understanding you correctly (with sensor placement aside) - to get the umol/joule number you're after we would perform this same test in a 1m2 (11.125" radius) integrating sphere. Correct? Now the next question I have is whether you'd like to assist in designing one?

Click to expand...

Yes, if you wanted to use that sensor without using any conversion multiplier, you'd want the innersphere surface area to be 1m2. You'd also want the sensor at a 90° placement along circumference from the testing source. You'd want a baffle to stop direct like to the sensor. You'd want to match your readings with the spectral response curve of the sensor. And you'd want to try to eliminate all light coming in or out.

Haha would I like to build one? Idk I'm thinking of making one myself but a quick search and I've found a link that shows a project thats already designed, I think you'd just need to print it out..
https://www.thingiverse.com/thing:554161

I think you can build them to be pretty accurate anymore, but a 3rd party confirmation goes a long ways and is ultimately what the masses will want..

HydroGrowLED said:

The sphere was custom engineered and 3D printed here at our offices. It can be used with a Light Scout or Li-Cor light meter.

Click to expand...

how are you calibrating this setup? anybody can take a random spot readings in a sphere. ive been doing it for years (and am always careful to present the data as relative to other data in the same test - its not a umol/J measure)

@CobKits can tell you more and in better detail, as he has actual experience using an integrating sphere, and is a better resource than I on the matter. Good luck.

OneHitDone said:

Here you go, just send a fixture to the lab used for these tests and if it comes back with respectable #'s all of the LED fanboys will have no choice but to eat crow

https://chilledgrowlights.com/independent-lab-reports

Click to expand...

Called them today, will find out tomorrow about their test and equipment.

ChiefRunningPhist said:

That's correct if shining onto a planar surface from a distance. That's not what's happening in a sphere though. The sphere acts like a big optic or lens. If its light proof then the light stays within the sphere bouncing all around. Because its a sphere it bounces the light equally all over, or as close to 100% as possible. Then when you take a point reading you can translate that over the entire sphere surface area to get a total μmol reading.

2 things. 1 is that if the sensor isn't picking up the ambient, how accurate is the sensor? 2, if you're losing light while you're testing, ie you can see light, your results will be less than what they should be.

The ability to detect μmol of the individual WVs will be based on the spectral response curve of the sensor you're using, which I haven't been able to track down yet. So some WV may be picked up easier or stronger then others, which would screw with the results too.

Yes, if you wanted to use that sensor without using any conversion multiplier, you'd want the innersphere surface area to be 1m2. You'd also want the sensor at a 90° placement along circumference from the testing source. You'd want a baffle to stop direct like to the sensor. You'd want to match your readings with the spectral response curve of the sensor. And you'd want to try to eliminate all light coming in or out.

Haha would I like to build one? Idk I'm thinking of making one myself but a quick search and I've found a link that shows a project thats already designed, I think you'd just need to print it out..
https://www.thingiverse.com/thing:554161

I think you can build them to be pretty accurate anymore, but a 3rd party confirmation goes a long ways and is ultimately what the masses will want..

Click to expand...

I do enjoy talking tech with you. Trying to understand your way of thinking is important to me. I do hear what you're saying about the light shining through the sphere. I thought it was understood that the readings were relative to my sphere and distance (as I had asked you in prior threads at what distance a umol/joule figure is calculated at since PAR readings change), but see now where my errors were in stating "per watt" which was only meant to show how much power the LED was using at the time of the reading (again relative). I agree that a third-party test is always great to backup in-house tests. I've called a company and will hear back from them tomorrow on that. I am curious now about the umol/joule calculation, but only by then applying it to my own spectral math equation.

I design everything in 3D myself. SOL, X-PRO, Alpha and X3 were all engineered by me. Every nut, bolt, spacer, fan blade, heat sink fin, etc... is all my CAD work. So designing things like an integrating sphere is fun for me, but I'd need to get some bigger printers to do a 1m2 sphere.

The problem with a sole μmol/j measurement, is that it's weighted red, just like the lumen/watt is weighted green. So in the end even a μmol/j measurement doesn't equate to a total "effeciency" metric. This is what I was saying earlier about "relative effeciency" being a more accurate measurement of effeciency.

It takes 18w to make 100μmol/s of 660nm photons. (Assuming 100% effeciency)
It takes 27w to make 100μmol/s of 440nm photons. (Assuming 100% effeciency)

^^^both WV have same photon count or μmol of photons but the energy used to achieve the photon count is different.

Let's say you need 1000μmol/s of photons. Let's say you have 2 options to achieve that photon count, either 440nm or 660nm. Lets say the 440nm chip is 60% effecient and the 660nm chip is only 39% effecient.

440nm:
1000μmol/s ÷ (100μmol/27J)
=
270w
.........
270w ÷ (60%)
=
460w of 440nm to achieve 1000μmol/s
:
184w of waste
........

660nm:
1000μmol/s ÷ (100μmol/18J)
=
180w
.........
180w ÷ (39%)
=
460w of 660nm to achieve 1000μmol/s
:
280w of waste
.........

Both have same μmol/j, but 1 is much more inefficient and requires much more thermal mgmt. If you duplicated an HPS SPD with an LED array, you could then compare μmol/j between the luminaries to determine cost savings at the meter, otherwise, if the SPDs are not the same, a μmol/j comparison is not indicative. An ineffecienct heavy red SPD can achieve the same μmol/j as a more effecient heavy blue SPD. 2 plants under the 2 lights could have drastically different growth responses but be under the same μmol/j.

What I was saying earlier was that relative effeciency is a more important metric than photon count per watt (I'm defining "relative effeciency" as how many photons are created, vs how many could be created), but I'm not sure that's true or if I believe that anymore. Even with electrical effeciency or "relative effeciency", it means nothing in a horticultural sense without knowing the SPD. Same with μmol/j. Same with lumen/watt. Like you said the only metric that matters is gram/watt. Once the optimal SPD has been determined, and if all lights are implementing such SPD, then any of the metrics can be used to compare effeciency with respect to yield or operating cost, ect, but till then, hypothesized gram/watt comparisons of luminaries will have to be a mesh of the SPD & μmol/j or relative effeciency, at least for me...

...
Theres a way to weight the μmol/j of an SPD to the mcree curve or peak plant absorption ranges and then multiply by relative effeciency.. or something ... but my math is rusty and I haven't deduced it yet lol ... but I'm sure someone somewhere has and that's ultimately the metric I think that should be a standard for horticulture. A μmol/j measurement weighted toward plant response with respect to the different energies of the different WV all multiplied by the average relative effeciency.. Or something...

ChiefRunningPhist said:

@CobKits can tell you more and in better detail, as he has actual experience using an integrating sphere, and is a better resource than I on the matter. Good luck.

Click to expand...

ive used a sphere, like OP, without calibration

so in my case useful for relative purposes. in OPs case, useful for marketing purposes

HydroGrowLED said:

I do enjoy talking tech with you. Trying to understand your way of thinking is important to me. I do hear what you're saying about the light shining through the sphere. I thought it was understood that the readings were relative to my sphere and distance (as I had asked you in prior threads at what distance a umol/joule figure is calculated at since PAR readings change), but see now where my errors were in stating "per watt" which was only meant to show how much power the LED was using at the time of the reading (again relative). I agree that a third-party test is always great to backup in-house tests. I've called a company and will hear back from them tomorrow on that. I am curious now about the umol/joule calculation, but only by then applying it to my own spectral math equation.

Click to expand...

what you need is a basic understanding of what a sphere is and how it works

there is no "distance at which you make a umol/J calculation". Thats like judging gas mileage by the speed on your speedometer

Snip from above...
...
Theres a way to weight the μmol/j of an SPD to the mcree curve or peak plant absorption ranges and then multiply by relative effeciency.. or something ... but my math is rusty and I haven't deduced it yet lol ... but I'm sure someone somewhere has and that's ultimately the metric I think that should be a standard for horticulture. A μmol/j measurement weighted toward plant response with respect to the different energies of the different WV all multiplied by the average relative effeciency.. Or something...

I thought @welight did something similar with the solskins?

To me this is a communication problem, in and amongst and between light and the members of this forum.

Light knows if it is particle or wave, we do not. Without taking into account generation losses and a full thermodynamic evaluation of the generating equipment and the ancillary systems it is pointless to have a discussion on effeciency. The only thing that can be conclusivly argued by a side by side is efficacy. Even so any result is subjective to the user so these giveaways are just a thinly veiled advertising tactic.

sethimus said:

he cut off his wiener, that should have been your first clue...

Click to expand...

Some might argue that to be a sign of sanity. Not me, but open your mind a smidgen sweetie, latent heat and latent homosexuality can lead to issues.

Maybe HGL should send the same chips to Teknik if he would be willing to test, so the the data could be compared.

CobKits said:

ive used a sphere, like OP, without calibration

so in my case useful for relative purposes. in OPs case, useful for marketing purposes

Click to expand...

In my case it was done for scientific purposes. After I learned that I could not count on a LED with higher lumen output producing higher umol output, I realized that I needed a new testing apparatus to compare LEDs from brands all over the world. The first thing my sphere was used for was to compare relative output from one brand and package to the next, to be sure that the LEDs we were using had the highest umol output per watt consumed.

After we selected the LEDs with the highest umol/watt for the 5 bands that fall within the PAR range, that's when we used the relative output from each wavelength to calculate our spectral ratios.

The information is also useful for marketing, but ultimately the goal is educating people at the same time.

CobKits said:

there is no "distance at which you make a umol/J calculation". Thats like judging gas mileage by the speed on your speedometer

Click to expand...

Yes, that has now been explained. I now understand the umol/J calculation thanks to @ChiefRunningPhist who has been most helpful in explaining it.

OK........Maybe we have bitten off more than we can chew here guys.......Apparently, the LEDs that Hydrogrowled are using must be tapping into zero-point energy. 54 umol/w how do you compete with that?!

Stephenj37826 said:

OK........Maybe we have bitten off more than we can chew here guys.......Apparently, the LEDs that Hydrogrowled are using must be tapping into zero-point energy. 54 umol/w how do you compete with that?!

Click to expand...

By doing zero innovation and slapping generic white LEDs on a board

I don't know what is truth or who to believe about what but I will say that @Stephenj37826 better tag @robincnn to take over cause @HydroGrowLED just slapped their strap on all up side your face with that fancy photosynthesis mathematical equation in the other thread and your in over your head

OneHitDone said:

I don't know what is truth or who to believe about what but I will say that @Stephenj37826 better tag @robincnn to take over cause @HydroGrowLED just slapped their strap on all up side your face with that fancy photosynthesis mathematical equation in the other thread and your in over your head

Click to expand...

The mathematical equations are not valid until an efficiency test is done. They really mean nothing until you know the amount of photons that the light is emitting overall. You will see once an efficiency test is done that HLG will be in front simply because of the diodes used are of a higher efficiency.

TEKNIK said:

The mathematical equations are not valid until an efficiency test is done. They really mean nothing until you know the amount of photons that the light is emitting overall. You will see once an efficiency test is done that HLG will be in front simply because of the diodes used are of a higher efficiency.

Click to expand...

I know you have the machines and all but I would disagree potentially. The light I used in the party cup comp has epistar reds and it will clean the clock of the cobs I have used previously and I feel confident saying qb's of comparable wattage as well.

I'm discussing the efficiency and maths behind the calculations. Regarding the spectrum for plant growth some grow very well under heavy reds, I have seen it several times but then not do so good at other times. I know the efficiency of the reds that are being used is not very high although if they were good chips they would destroy a HLG efficiency because they are heavy reds.