Lets talk about "Other" COBS !

[PurpleBuz]

Will test PAR of each COB all driven off the same driver in the same environment,
I will test them without optics and then with the 80 degree Glass Optics.
This will be over a 5 x 5 space and check the numbers and see if we came close or lost it. Things are looking good.

A ppfd footprint over a 5x5 has too many variables to estimate PPF. Much more interested in raw PPF under controlled consistent conditions. For comparing different cobs, testing with\without optics is also less interesting.

I assume you don't have an integrating sphere, but if you setup a white reflective tube about 12" long and maybe 6 inches diameter with the cob at one end, and a decent par meter at the other end. While this can't be compared directly to PPF from an integrating sphere, it should be good enough to compare cobs measured the same way, and provide a relative comparison to the PPF in the datasheets.

 

[shaft09]

I don't know much about LED's and drivers yet, but ran across this

http://www.trpssl.com/led_driver_selector.html

http://www.trpssl.com/light-engines_round.html

 

[coolj]

this is good Sh**!!!!

 

[robincnn]

I assume you don't have an integrating sphere, but if you setup a white reflective tube about 12" long and maybe 6 inches diameter with the cob at one end, and a decent par meter at the other end.

For cob comparison testing without optics... do we need a tube or sphere?
All cobs have same radiation pattern... lambertian
Checked 3 diffrent cob companies ..all lambertian
Differences in LES size can be ignored if you are measuring at 12 inches and Les is smaller than 1 inch
What do you think of just center measurements at a fixed distance from cobs in a open area
I think that is how Supra made the cob comparison charts

 

[PurpleBuz]

For cob comparison testing without optics... do we need a tube or sphere?
All cobs have same radiation pattern... lambertian
Checked 3 diffrent cob companies ..all lambertian
Differences in LES size can be ignored if you are measuring at 12 inches and Les is smaller than 1 inch
What do you think of just center measurements at a fixed distance from cobs in a open area
I think that is how Supra made the cob comparison charts

I thought a tube like a 6" PVC would be good to eliminate variances in reflection, that one would get in an open area or a tent. but really not needed.

 

[BOBBY_G]

Watching this with interest, it would most certainly end the HPS COB cost debate.
This with some of the heat sink seeking that is taking place could change the game completely, so keep it coming.

im not sure why there is a debate. even with current 3590 prices, you make it up in electricity and bulb costs within 2-3 years.

cheap cobs dont make the drivers or heatsinks any cheaper

 

[JorgeGonzales]

For cob comparison testing without optics... do we need a tube or sphere?
All cobs have same radiation pattern... lambertian
Checked 3 diffrent cob companies ..all lambertian
Differences in LES size can be ignored if you are measuring at 12 inches and Les is smaller than 1 inch
What do you think of just center measurements at a fixed distance from cobs in a open area
I think that is how Supra made the cob comparison charts

I think you are exactly right about the lambertian curve, as long as the measuring distance is fixed it should provide good relative comparisons between cobs. I'll take your word on the 12 inch/1 inch LES math.

Isn't that how Supra measured everything? A lux meter while watching temps and watts/Vf to verify and extrapolate past what Cree provided in their data sheets.

 

[BuddyColas]

For cob comparison testing without optics... do we need a tube or sphere?
All cobs have same radiation pattern... lambertian
Checked 3 diffrent cob companies ..all lambertian
Differences in LES size can be ignored if you are measuring at 12 inches and Les is smaller than 1 inch
What do you think of just center measurements at a fixed distance from cobs in a open area
I think that is how Supra made the cob comparison charts

Wait a minute. You mean comparing cobs at like wattages then measuring 12" below the LES with a like PAR meter? Then comparing umols to dissapation watts between the cobs? Oh, it just has to be more complicated than that.

 

[robincnn]

@JorgeGonzales yes i think that is how supra measured it. I woulkd like to think he used a apogee meter and not lux meter

@BuddyColas Since the pattern of cobs is same. it will work. you get a ratio of umols PPFD directly under cob for specific watts for each cob.. get different data points and plot away... assuming PAR meter takes care of spectrum differences.. still saving for that new apogee meter....You can make a simple problem as complicated as you would like. This is not a mars landing mission and a little bit of errors don't leave me on the wrong planet. Cheep slightly inaccurate data better than no data. It is better than comparing lm/w from datasheets of different manufacturer... they all have different corporate guidelines of massaging numbers

 

[BuddyColas]

@JorgeGonzales yes i think that is how supra measured it. I woulkd like to think he used a apogee meter and not lux meter

@BuddyColas Since the pattern of cobs is same. it will work. you get a ratio of umols PPFD directly under cob for specific watts for each cob.. get different data points and plot away... assuming PAR meter takes care of spectrum differences.. still saving for that new apogee meter....You can make a simple problem as complicated as you would like. This is not a mars landing mission and a little bit of errors don't leave me on the wrong planet. Cheep slightly inaccurate data better than no data. It is better than comparing lm/w from datasheets of different manufacturer.

Robincnn, I totally agree. I was being a smartass...a bit of sarcasm I guess to all the nitpicking and hand-wringing and pages and pages of posts that goes on here about testing and comparing and needing a sphere to make a definitive decision between cobs and all. I think something as simple as you have suggested would be a great way to compare cobs. Very practical and 12" above the canopy is often a height we actually deploy them. I admire your work...please carry on and help the cob community. I am anxious to see your results. Thanks.

 

[PurpleBuz]

It is better than comparing lm/w from datasheets of different manufacturer... they all have different corporate guidelines of massaging numbers

very well put.

 

[JorgeGonzales]

Robincnn, I totally agree. I was being a smartass...a bit of sarcasm I guess to all the nitpicking and hand-wringing and pages and pages of posts that goes on here about testing and comparing and needing a sphere to make a definitive decision between cobs and all. I think something as simple as you have suggested would be a great way to compare cobs. Very practical and actually how we deploy them. I admire your work...please carry on and help the cob community. Thanks.

Still, an integrating sphere would be nice. I really hope somebody gets a cob light in one someday. Not to compare cobs, but to compare Supra's numbers with reality and make sure all this mickey mouse math is in the ballpark. Benchmarking against companies like BML and LSG would certainly be damned interesting too.

 

[Airwalker16]

im not sure why there is a debate. even with current 3590 prices, you make it up in electricity and bulb costs within 2-3 years.

cheap cobs dont make the drivers or heatsinks any cheaper

Because cxb's are $30-$40. These are $3-$6. DUH.
10 cobs for $400 or for $60? Hmmmmm....

 

[BOBBY_G]

Because cxb's are $30-$40. These are $3-$6. DUH.
10 cobs for $400 or for $60? Hmmmmm....

from the first examples we looked at, you need 3 of the $8 cobs to equal the efficiency of the $30-$40 cobs with the presumably long lifespan and lumen maintenance. as well as having to solder them up and then do 3x the drilling and tapping. im sure there are some better deals out there than cree if we do our homework, but its not as dramatic as replacing a cob with something 1/10 of the price and expecting to get the same results

 

[Airwalker16]

from the first examples we looked at, you need 3 of the $8 cobs to equal the efficiency of the $30-$40 cobs with the presumably long lifespan and lumen maintenance. as well as having to solder them up and then do 3x the drilling and tapping. im sure there are some better deals out there than cree if we do our homework, but its not as dramatic as replacing a cob with something 1/10 of the price and expecting to get the same results

You don't need 3 of them to get the same power/lumens as a cxb. .

 

[PurpleBuz]

but to compare Supra's numbers with reality and make sure all this mickey mouse math is in the ballpark.

wow dude show a little respect.

and btw the OP is the one that's gonna do some testing, so maybe ask the OP if he has an integrating sphere?
we were not discussing "Theory of Testing" I was making suggestions to the OP so that he could get repeatable consistent test results without the expense of an integrating sphere.

 

[JorgeGonzales]

wow dude show a little respect.

and btw the OP is the one that's gonna do some testing, so maybe ask the OP if he has an integrating sphere?
we were not discussing "Theory of Testing" I was making suggestions to the OP so that he could get repeatable consistent test results without the expense of an integrating sphere.

I literally didn't even notice you in this thread, but way to find a fight when there isn't one. You really just like starting shit where there isn't anything to start.

Feel free to troll through my comments and count how many times I say I've relied one on Supra's spreadsheets...oh look there I am saying it to you in another bullshit argument you started where you were completely wrong in your "calcs" while proclaiming black to be white with complete authority.

Mickey mouse math wasn't an insult. The reality is we are at the mercy of incomplete manufacturer data, extrapolating data points from incomplete graphs, digitizing SPD graphs, and in general hoping that if you make enough guesses and they all agree they are probably in the ballpark.

I'm sorry I'm being grumpy, but honest to god I was digitizing Nichia efficiency vs current graphs in another window, for this thread, when I open this up and find another giant bunch of horseshit waiting.

 

[BOBBY_G]

You don't need 3 of them to get the same power/lumens as a cxb. .

ok heres my math, lets see your take on it

https://www.rollitup.org/t/lets-talk-about-other-cobs.908494/page-3#post-12578556

 

[JorgeGonzales]

It is better than comparing lm/w from datasheets of different manufacturer... they all have different corporate guidelines of massaging numbers

If you want to see some number massaging, I just digitized a couple charts from the Nichia NFDLJ130B datasheet, to get a rough idea how it behaved. @BuddyColas asked for some extra some eyeballs on it awhile back.

Nichia gives a wide range of lumen output, and all the charts are at Ta=25C, which I'm guessing is a pulsed test in a temperture controlled environment, so probably the same as Tj=25C? Anyway:

NFDLJ130B 3500K CRI 80 Ta=25C
6210 - 7610 lumens @1150ma 39.2 Vf 45.08W 138-169 LPW
3273 - 4010 lumens @545ma 36.2Vf 21.27W 154-189 LPW

If the middle of the range is "typical" then it keeps up with a Vero 29 at Tj=25, and I think Buddy said it was only $5? That's pretty wild.

He also mentioned the NFCLJ108B which I'll try to look at next, I think that was the better performing one for $7.

 

[JorgeGonzales]

ok heres my math, lets see your take on it

https://www.rollitup.org/t/lets-talk-about-other-cobs.908494/page-3#post-12578556

Honestly, once you start hitting the big Cree numbers at 180-200lm+ per watt, it's hard for most cobs to match them at any current.