Top bin COB comparison 2

[Rocket Soul]

After lurking and reading on the forum for somewhile back i was so happy to find theses bin comp threads with real measured data. I was starting to doubt some of the spreadsheet maths people use when i saw the readings people got from one of the optic/reflector threads (https://www.rollitup.org/t/lens-and-reflector-optics-for-cob.893660/), result just didnt make sense with regards to some calculations ive seen based on the spreadshhets. So real measured data was like yes finally! Hats off to supra and robin cnn and everyone else with the will and equipment to do this. But the graph that supra put up on the first bin comp thread got me thinking in the end: consistently higher ppfd/watt@x-current the higher K the spectrum was, comparing the same model for example the cxb3590. Now that doesnt make no sense to me. Please dont take this as knowitall-ism, i know very well im not as versed in the math nor the science in all this but there is something fishy in this: we know that blue light is higher in energy/photon, meaning that for 1 watt of light, the more red the more photons, although each with less energy. If we are measuring this correctly, and the conversion bit based upon the apogee not reading all the red photons, shouldnt it be the other way around? Ok, so the bluer cobs have higher bin, and higher lumen. But all the wattages is the same for every measurement so shouldnt blue come out losing in ppfd/watt@x-current? The differnce between 1 watt of blue and 1 watt of red or far red was stated around 50ish % more photons in the red spectrum. I mean this is the whole reasoning for reder lights for growing as photosynthesis is driven by amount of photons? Not sure if im clear enough, but theres something that doesnt make sense with those graphs. Not sure if its in the maths(dont think so, supra knows his shit), in the apogee readings (how do they work anyway, they actually count photons? Or aproximate how many based on the watts of radiant energy they recieve and compare it to which color? Anyone knows?) or maybe the conversion factor (most likely?). Please, anyone has an idea?

 

[Rocket Soul]

Anyone tried the following test: comparing heat disipation on different color spectrum of cxb each under same conditions? If the fixture efficiency loss due to heat are all the same then lower spectrum temprature must have higher ppfd as each photon would spend less of the radiant/par watts. If the graph is correct then higher spectrum temprature should be dramaticly easier to cool as not only are they shootin of more photons per watt, the bluer photons also carry more energy each one, leaving very little energy left from total disipation wattage for the heatsink to absorb.

 

[BOBBY_G]

Anyone tried the following test: comparing heat disipation on different color spectrum of cxb each under same conditions? If the fixture efficiency loss due to heat are all the same then lower spectrum temprature must have higher ppfd as each photon would spend less of the radiant/par watts. If the graph is correct then higher spectrum temprature should be dramaticly easier to cool as not only are they shootin of more photons per watt, the bluer photons also carry more energy each one, leaving very little energy left from total disipation wattage for the heatsink to absorb.

theyre not dramatically different to cool. were talking 15% efficiency difference from CB to DB bin

 

[Rocket Soul]

Hey bobby g. Not sure if im barking up the wrong tree with this. Still havent figured out how to qoute and how to bring the graph into this post, not posted more than once or twice. Maybe better explaining where i started my train of thought. I was looking at supras graph and thinking about flowering. The 5000k db line comes in as highest ppfd/watt at every current. What does this tell me, everything being correct? Well if flowering under bluer light generally would give you more potent bud, and if the blue db gives you higher ppfd/watt, then the best flower light would be a 5000k db, as it gives more photons (higher photosynthesis as its driven by nr of photons) and higher potency as its more blue. That kinda goes against common sense and experience of most people growing? As in more yield with red, higjer potency with blue. And if this is true the disipation watts should be quite skewed towards light instead of heat. Maybe the 15% you quote is just this. What im saying is that maybe theres another way at getting at these nrs, to see if theyre actually good: instead of measureing ppfd and charting if against watts, is there a way to figure this out thru the heat dissipation as what doesnt become light has to become heat. This could of course only be done if the bins tested for 5000k and 3000k where the same. And i guess plant response maybe also is not linear in proportion to amount of photons. But thengraph isnstill saying 5000k db has more photons and more enrgy per photons.

 

[Rocket Soul]

The thing that maybe makes this tricky to grasp is when we see these graphs we tend to think that 3000k would be more efficient since the warm spectrum has more photons per lumen, but these graphs allready correct for that as we are measureing ppfd/watt, not lumens per watt. Maybe your right in that this is for the 5000k having higher bin, but then this data says for highest photosyntesis and highest potency, go for cold intense luminosity in flower which doesnt seem to sit right with what most people say/do. @Supra, would you have a look and weigh in?

Edit for spelling

 

[BOBBY_G]

Hey bobby g. Not sure if im barking up the wrong tree with this. Still havent figured out how to qoute and how to bring the graph into this post, not posted more than once or twice. Maybe better explaining where i started my train of thought. I was looking at supras graph and thinking about flowering. The 5000k db line comes in as highest ppfd/watt at every current. What does this tell me, everything being correct? Well if flowering under bluer light generally would give you more potent bud, and if the blue db gives you higher ppfd/watt, then the best flower light would be a 5000k db, as it gives more photons (higher photosynthesis as its driven by nr of photons) and higher potency as its more blue. That kinda goes against common sense and experience of most people growing? As in more yield with red, higjer potency with blue. And if this is true the disipation watts should be quite skewed towards light instead of heat. Maybe the 15% you quote is just this. What im saying is that maybe theres another way at getting at these nrs, to see if theyre actually good: instead of measureing ppfd and charting if against watts, is there a way to figure this out thru the heat dissipation as what doesnt become light has to become heat. This could of course only be done if the bins tested for 5000k and 3000k where the same. And i guess plant response maybe also is not linear in proportion to amount of photons. But thengraph isnstill saying 5000k db has more photons and more enrgy per photons.

while photons are photons to an extent flowering plants can utilize red light more efficiently. so a 3000K cob might outperform a brighter 5000k cob for flowering
im no scientist but i would venture to guess the difference in weight from adding red i s far more dramatic than the difference in potency by adding blue

 

[Rocket Soul]

I was gonna answer yesterday but my couch locked on me...
Your right, but i think i made a bad example bringing the flowering into this, im talking about the numbers, not photosynresis. If redest part of the spectrum should carry around 50% more photons per watt than the bluest part then why does blue come out so on top in ppfd which is photons/area/second. I would expect the warmer lights to do a bit better against the blue. In this graph, where supra did some more carefull and repeated measurements we have a nice comparison: 5000k cd2 versus 3500k cd2,both of the same bin so same lumens and coming out very closely. But the again the colder light comes out higher ppfd, even if in theory testing in this way should correct for the color temp. Also, spreadsheets say they dispate heat at same efficiencies.

At same lumens colder light should have a fair bit less photons, watt per watt, even though these photons carry more energy each. So i maintain something is wrong: either the apogee is biased in its reading depending on color temp, well or the conversion factor is wrong, or physics are wrong. Well actually we know the apogee is biased, not reading all the reds, so the conversion factor isnt acurate. I dont doubt supras maths or measureing, but somewhere something has snuck in to throw the result.

 

[Greengenes707]

Stoke loss
The fact that all top bin use the same die...higher k=less phosphor=less stoke loses to the base performance of the die.

If you were comparing the same energy then yes, a red spectrum has a higher potential(LER). But that isn't what is going on.

 

[Rocket Soul]

Stoke loss
The fact that all top bin use the same die...higher k=less phosphor=less stoke loses to the base performance of the die.

If you were comparing the same energy then yes, a red spectrum has a higher potential(LER). But that isn't what is going on.

Hi Gg. I left this thread to have a think before getting myslef into trouble. Quote"comparing at the same energy then yes, red spectrum has a higher potential(LER)" ive not really got my head around LER, but the whole point with these graphs are that we are comparing at same energy, for every point on the x-axis we are using the same watts. I might be wrong but wouldnt that mean the warmer light source, at same bin and power should come out on top, having more red spectrum?

 

[Rocket Soul]

Is it possible that 3500k cd bin and 5000k cd bin doesnt use the same die? 3,5 cd and 5k db seens to be the same family, same die but not the same phosphorous, but 5k cd could be another bin generation. Idk, just trying to make sense out of this, please set me right if im wrong

 

[robincnn]

Y Axis = µmol m−2s−1 at 14"
X-Axis = Pulsed 25watts/50watts/75watts
Notes: All cobs 80 CRI. No spectrum correction
SQ 120


@BOBBY_G you have a better setup. Sending Vero 29 and 1818 3000K samples for more testing.

Edit: This is type D vero

 

[Malocan]

View attachment 3718815


View attachment 3718804
Y Axis = µmol m−2s−1 at 14"
X-Axis = Pulsed 25watts/50watts/75watts
Notes: All cobs 80 CRI. No spectrum correction
SQ 120
View attachment 3718810

@BOBBY_G you have a better setup. Sending Vero 29 and 1818 3000K samples for more testing.

nice work Robincnn, Thanks for sharing

 

[JorgeGonzales]

View attachment 3718815


View attachment 3718804
Y Axis = µmol m−2s−1 at 14"
X-Axis = Pulsed 25watts/50watts/75watts
Notes: All cobs 80 CRI. No spectrum correction
SQ 120
View attachment 3718810

@BOBBY_G you have a better setup. Sending Vero 29 and 1818 3000K samples for more testing.

This is great.

Looks like the 1818 and Vero 29 might meet in the middle if their CCTs were the same. Did not expect that.

I am suspect of the differences between 4000K and 3000K, though. Think that's a sensor bandwidth/curve anomaly.

And why is the Cree suddenly at the bottom of all of these charts?

 

[robincnn]

And why is the Cree suddenly at the bottom of all of these charts?

Cree old tech and expensive.
Even with the bin jump the gap may be narrow between 3590 and Vero 29.
Cree just has better voltage compatibility with drivers.

Brdigelux has reduced voltage in new cob. For 50 watts 36.65V compared to 37.6

@BOBBY_G has a better sensor so lets see what are his results with the new Vero 29D and 3590CD

 

[AquariusPanta]

While new to the Citizen cob scene, I'm happy to see the shift of focus from the more prevalent brands here in the forum; New is always good and interesting!

In the near future, when more data is collected on the above models that robincnn charted, would it not be more appropriate we compare Vero 29 type B with the 1818's or should we compare with 1825's given similar forward voltage? Type B and C makes appear more efficient than their type D counterpart.

Concerning Cree, I wouldn't be surprised if they retake the hill of efficiency by the end of the year. CXC would have to be a game changer and the game ain't getting any easier (or maybe it is? hah).

 

[JorgeGonzales]

While new to the Citizen cob scene, I'm happy to see the shift of focus from the more prevalent brands here in the forum; New is always good and interesting!

In the near future, when more data is collected on the above models that robincnn charted, would it not be more appropriate we compare Vero 29 type B with the 1818's or should we compare with 1825's given similar forward voltage? Type B and C makes appear more efficient than their type D counterpart.

Concerning Cree, I wouldn't be surprised if they retake the hill of efficiency by the end of the year. CXC would have to be a game changer and the game ain't getting any easier (or maybe it is? hah).

I'd say it makes the most sense to compare price points. Without Bridgelux prices, that is difficult to do right now. An 1825 has the potential to be much cheaper than anybody is selling CXB3590s for, once they are stocked by whoever plans on stocking them. Same with the 1212, and 1818. Cheaper than any Vero 18, or Vero 29 with near certain odds, at least at Digikey prices.

Vero 29 sure looks nice though.

 

[AquariusPanta]

I'd say it makes the most sense to compare price points. Without Bridgelux prices, that is difficult to do right now. An 1825 has the potential to be much cheaper than anybody is selling CXB3590s for, once they are stocked by whoever plans on stocking them. Same with the 1212, and 1818. Cheaper than any Vero 18, or Vero 29 with near certain odds, at least at Digikey prices.

Vero 29 sure looks nice though.

Are we basing data off of version 5 or unreleased version CLU cobs? Again, I'm late to the show

 

[JorgeGonzales]

Are we basing data off of version 5 or unreleased version CLU cobs? Again, I'm late to the show

Version 5, they are relatively recent. Same generation as gen7 Bridgelux, basically.

 

[AquariusPanta]

Version 5, they are relatively recent. Same generation as gen7 Bridgelux, basically.

I don't know if Digikey is a good source to base prices on, as there may likely be cheaper sources if not already, but they do have some of the 1212 (~28$ per x1) in stock, with 1818 prices available (~53$ per x1) but with no inventory.

 

[JorgeGonzales]

I don't know if Digikey is a good source to base prices on, as there may likely be cheaper sources if not already, but they do have some of the 1212 (~28$ per x1) in stock, with 1818 prices available (~53$ per x1) but with no inventory.

CDI sometimes has stock, but a 1212 is $12 there. This is a much more realistic price. Digikey is high.