Spectral lightmeter Gigahertz-Optik MSC15

[JorgeGonzales]

You know, the more I think about it...the less min/max seems like a good way to measure uniformity. Easy to calculate maybe, but not particularly accurate. In Crees case, one dim module would throw off the whole calculation, in @Malocan's case, a slight tilt keeps the center max the same, but has a dramatic effect on the edges, fucking the numbers.

Calculating the mean, and summing distance from the mean or maybe just mean deviation? This gives equal weight to each measurement, not just the maximum and minimum.

 

[guod]

uniformity is the key...
where less light, there is less growth

 

[JorgeGonzales]

 

[Johnnycannaseed1]

You know, the more I think about it...the less min/max seems like a good way to measure uniformity. Easy to calculate maybe, but not particularly accurate. In Crees case, one dim module would throw off the whole calculation, in @Malocan's case, a slight tilt keeps the center max the same, but has a dramatic effect on the edges, fucking the numbers.

Calculating the mean, and summing distance from the mean or maybe just mean deviation? This gives equal weight to each measurement, not just the maximum and minimum.

Agreed...

If I could make a suggestion, I recommend after you have worked out the mean, that you take that figure and use it to calculate the variance of the numbers.

In this case a low variance = The Values in the number set are close to one another... = Good uniformity

Whereas a high variance = The Values in the number set are more spread apart... = Poor uniformity

Now you can compare different fixtures uniformity based upon Variance Value.

See link below on how to calculate Variance

http://www.wikihow.com/Calculate-Variance

Seeing as ft/sq is one of the standard measurements in horticulture, might be worth the light manufacturers in here calcualting Variance based upon 5 measurements per foot square (4 corners 1 on centre), then scaling it up over the area you are measuring.

Example 16 Sq/ft = 41 datapoints... then work out the mean value... then calulate the Variance.

That way you could have a standard whereby you can do comparissons of any grow light fixture LED or otherwise by calculating the uniformity of the fixture over a given area... Just a thought mind

 

[Johnnycannaseed1]

uniformity is the key...
where less light, there is less growth

Uniformity is definitely key but equally so is the correct amount of light...

To little as you have correctly pointed out = less growth.

But conversely high light levels can equal a waste of energy (best case scenario) or light saturation/photoinhibition worst case (unless running Co2)... Hence growers with overpowered fixtures having to hang them further away = waste of energy.

Finely balancing all the variables together is the way forward...

 

[JorgeGonzales]

@Malocan, you might want to try lowering the no-reflector fixture to see if you can match the average PPFD of the reflector version, with greater uniformity, or at least see if that goal is possible.

This thread is gold.

 

[nevergoodenuf]

@Malocan I found you a new toy, so I can borrow that old, outdated, piece of junk you are currently using.
http://www.ledsmagazine.com/ugc/2016/08/05/xitrons-portable-microspectrometer-offers-lab-instrument-performance-in-led-lighting-applications.html

 

[Rocket Soul]

Subbed

 

[Malocan]

@Malocan, you might want to try lowering the no-reflector fixture to see if you can match the average PPFD of the reflector version, with greater uniformity, or at least see if that goal is possible.

This thread is gold.

Hello Jorge,
sorry for my late answer, but i had a lot stuff to do in real life.
And i cant answer your question because i dont have access to the tent with the no-reflector fixture at the moment. I moved the tent to my good friends home :/

But i will try to make a smillar kind of test, but this will take 2-3 days. I have a lot of stuff still too do

Since a week i got my other clu58(3000k and 5000k both 80 cri) chips which i ordered @tme.eu
I will share some tests again cxb3590(cd 3500k)

And i did rebuild my clu58 2700k 90 cri fixture, i will share some photos
i added a lot of osram 660 reds , way to much maybe
I will runs the osrams only around 150ma i think, but maybe more i will see in future

 

[The Dawg]

 

[wietefras]

i added a lot of osram 660 reds , way to much maybe
I will runs the osrams only around 150ma i think, but maybe more i will see in future

Did you also measure only the deep red Osrams without the COBs switched on? As an indication of how much watts of those you need to adjust a spectrum to where one would like it.

 

[salmonetin]

...maybe im wrong...seems to me... if we need adjust spectrum ...the measure must include the whites to show the mix... only the red measurement dont give us the correct info...for adjust the spectrum to where one could like it....

...but its easy... a comparative ...with only reds ... and with the whites involved.... and see the results...



Saludos

 

[wietefras]

He already posted the combo of white and red together above. He also posted the whites alone. So I was wondering about the reds alone.

Guess I could subtract the whit from the combo though.

 

[Malocan]

Did you also measure only the deep red Osrams without the COBs switched on? As an indication of how much watts of those you need to adjust a spectrum to where one would like it.

Hello,
the red osrams runs around 150-200ma on the first picture and one the 2nd picture they runs around 350-400ma.
The cob is on both pictures around 1ampere.



1st picture: 49watt cob + 6-9watt red (not exactly)
2nd pciture: 49att cob + 17-20watt red (not exactly)

i hope this helps

 

[JorgeGonzales]

I'm not saying this is an ideal spectrum or anything, but I think it's pretty funny that the Fluence Physiospec Indoor spectrum is obviously a white led with some 660nm reds added:



I actually digitized this, and once I started combining 660nm LEDs with 4000K white leds in Excel it got very similar very fast, enough to convince me that was all that's going on.

 

[wietefras]

i hope this helps

Sure does, thanks! I

I missed them earlier, but I love the footprint charts you made comparing bare COBs vs reflectors: http://rollitup.org/t/spectral-lightmeter-gigahertz-optik-msc15.913239/page-3#post-12730919

Good to see an actual real life situation tested and it conforms exactly to what I tried to explain months ago:

The chart shows percentages of extra light you get when you use bare COBs instead of reflectors from comparing the two charts. Overall you estimated 629 PPFD average for the COBs with reflectors @ 60cm and 744 PPFD for the bare COBs @ 31.5cm. Which are the heights where you would hang the fixture with or without reflectors to get proper uniformity. So you get 18% more light when used without reflectors!

I have to admit I didn't anticipate reflectors being that bad, but this is a reasonably big tent so the reflectors make even less sense here. Also I guess 61cm is a bit too high for 90 degree reflectors. Going from 115 degrees @ 31.5cm to 90 degrees @ 50cm would be more comparable, but that difference is small (say around 5%).

Still it's pretty much exactly what I said all along. What's even more funny is that I said, the only scenario where reflectors really work is when you don't have (reflective) walls. Low and behold, the only area where the reflectors give you more light on the plants is at the door opening ...

If you leave that first row out you get an average PPFD of 637 vs 764. Which is even a 20% difference in favor of bare COBs.

BTW Maybe you should try some tests with other reflective materials on the walls. I saw another test where diamond foil performed very poorly compared to more reflective surfaces including mirrors. Diamond foil has a retroflective pattern which means the light will get reflected back to the origin of the light. You could probably get better uniformity close to the walls with reflective foils or mirrors

 

[cdgmoney250]

Testing spread without reflective walls would be beneficial for anybody with a larger size grow. Large canopies don't have the luxury of having much/any wall reflectivity.

Reflectivity will skew stand alone results.
Just think larger than a tent.

 

[salmonetin]

Hello,
the red osrams runs around 150-200ma on the first picture and one the 2nd picture they runs around 350-400ma.
The cob is on both pictures around 1ampere.



1st picture: 49watt cob + 6-9watt red (not exactly)
2nd pciture: 49att cob + 17-20watt red (not exactly)

i hope this helps

...i don saw the pics...sorry ... maybe my mozilla...or my old pc...

Pd...i upload a doc zipped ...its a typical backup mine ...with the info on web pages i do it a doc...nothing too special... its for read on home with more time....enjoy



Saludos

 

[Malocan]

...i don saw the pics...sorry ... maybe my mozilla...or my old pc...

Pd...i upload a doc zipped ...its a typical backup mine ...with the info on web pages i do it a doc...nothing too special... its for read on home with more time....enjoy



Saludos

Thank you very much salmonetin,
i will read it before i going sleep, my sleep lecture A lot of informations, intersting

Thanks

 

[ReddEyez]

Great thread everybody! Thank you for all the testing and info. Great job Malocan, thank you for all your time! Your lights look great man!