Testing Timber's prototype=4000k/660nm and my finger on the flower trigger:)
- Thread starter PSUAGRO.
- Start date
BM9AGS
Well-Known Member
I'd like to see your results for sure. Simply because the results of the cobs is goooooood in real grows on real plants. This all makes me want to do a grow with red only blue only why're only and green only with clones. Would be interesting to see results. Hasn't this been done???
Last edited:
Stephenj37826
Well-Known Member
Yeah I'm wondering the same about temps as well. At those temps 212 watts makes sense. That combo should've been pulling around 220 but I guess with massive thermal droop maybe not so much......... I would still like to see 3rd party independent testing done where a competitor to the general tech isn't involved. Oh and don't think I'm being biased because our fixtures use SMD and COB. They both are great and in honesty work on the same exact principles just different packages. 150F ambient is acceptable for testing high bay or something but is not really "real world" in a growing environment. Furthermore I would like to see the temps that the Nextlight series was tested at as well. Was these tests truly single variable???????
Last edited:
robincnn
Well-Known Member
@PSUAGRO Thanks for the test data
. Very useful. First ever COB sphere data in this forum
Must say it was raw and needed to be cooked for some lazy people including me. I was too lazy to read old
65C=150F
Calculated Values
http://pct.cree.com/dt/index.html
COB Level 146lm/w ( assuming 3000K was CB)
COB Level 159.3lm/w ( assuming 4000K was CD)
Assuming HLG-185-1400 was used to test 4 cobs. Driver efficiency typical 93%.
So 146*0.93= 135.78 lm/W
and 159.3*0.93= 148.14 lm/W
Observed values
3000K lamp- 130.8 lm/w
4000K lamp- 135.7 lm/w
For 3000K that's like a difference of 3.8%. Within error margin. may be LES was not cleaned.lol . May be driver was not happy with the cobs.
We got to reduce calculated number just a little bit because i assume the reflector adapter was on, even though the reflector was not on ?. adapter would have some loss.
May be the 150F was ambient temp. Which means the Tc (Tj in calculator) should have been higher in calculator. may be 75 instead of 65. which means calculated values need to be lowered.
So for 3000K calculated values using tool are almost identical to what you found is sphere.
For 4000K i see a bigger difference. Assuming 4 cobs any small variations should have even out.
In observed values the 3000K and 4000K differ by only 5lm/w. It should have been way more, I mean 4000K is supposed to have higher Lm/w and it is 1 bin higher. Hard to interpret there
I think cree tool does give reasonably accurate data.
Supras sheet where off. I realized that when i tried to compare the Vero Gen 6 and Cree's and since then have been using cree PCT calculator instead.
Bro you used some outdated 2015 chips there.
2016 is about Vero Gen 7, Citizen. just kidding...
Thanks for taking the time and helping everyone.
EDIT: Calculated numbers also need to be reduced another 1lm/w because i think the driver was pushing 1450ma instead of 1400ma. at 1450ma 49.83*4/0.93=214.32
. Very useful. First ever COB sphere data in this forum Must say it was raw and needed to be cooked for some lazy people including me. I was too lazy to read old
65C=150F
Calculated Values
http://pct.cree.com/dt/index.html
COB Level 146lm/w ( assuming 3000K was CB)
COB Level 159.3lm/w ( assuming 4000K was CD)
Assuming HLG-185-1400 was used to test 4 cobs. Driver efficiency typical 93%.
So 146*0.93= 135.78 lm/W
and 159.3*0.93= 148.14 lm/W
Observed values
3000K lamp- 130.8 lm/w
4000K lamp- 135.7 lm/w
For 3000K that's like a difference of 3.8%. Within error margin. may be LES was not cleaned.lol . May be driver was not happy with the cobs.
We got to reduce calculated number just a little bit because i assume the reflector adapter was on, even though the reflector was not on ?. adapter would have some loss.
May be the 150F was ambient temp. Which means the Tc (Tj in calculator) should have been higher in calculator. may be 75 instead of 65. which means calculated values need to be lowered.
So for 3000K calculated values using tool are almost identical to what you found is sphere.
For 4000K i see a bigger difference. Assuming 4 cobs any small variations should have even out.
In observed values the 3000K and 4000K differ by only 5lm/w. It should have been way more, I mean 4000K is supposed to have higher Lm/w and it is 1 bin higher. Hard to interpret there
I think cree tool does give reasonably accurate data.
Supras sheet where off. I realized that when i tried to compare the Vero Gen 6 and Cree's and since then have been using cree PCT calculator instead.
Bro you used some outdated 2015 chips there.
2016 is about Vero Gen 7, Citizen. just kidding...
Thanks for taking the time and helping everyone.EDIT: Calculated numbers also need to be reduced another 1lm/w because i think the driver was pushing 1450ma instead of 1400ma. at 1450ma 49.83*4/0.93=214.32
Last edited:
CobKits
Well-Known Member
so what youre saying is..... the manufacturers calculators are actually useful tools for predicting chip performance.
PSUAGRO.
Well-Known Member
Listen fellas not trying to start some stupid war , just wanted ANY info on the very popular cxb line........ can't ask for multiple chips & mah readings, dozens of reflector / optics work on volunteer time..... sorry, someone else has to pony up.
I'm on the road and have other issues to worry about........ hopefully I can get some more answers after the weekend on the specific test parameters.
...... this is what happens when you plant too far from each other(bad row spacing) with incompetent off-season pruning and mother nature "blows on your balls"....... fuck
Robincnn is always on point without acting like a pompous know it all........... thanks for the time brother, I appreciate the help....we're getting more sphere testing soon supposedly, will see on accuracy..... thanks
I'm on the road and have other issues to worry about........ hopefully I can get some more answers after the weekend on the specific test parameters.
...... this is what happens when you plant too far from each other(bad row spacing) with incompetent off-season pruning and mother nature "blows on your balls"....... fuckRobincnn is always on point without acting like a pompous know it all........... thanks for the time brother, I appreciate the help....we're getting more sphere testing soon supposedly, will see on accuracy..... thanks
Maybe you didn't get the memo?...... don't want you in here and I like the taste of crow!
Last edited:
Stephenj37826
Well-Known Member
Definitely not trying to rake you over the coals on this. Getting sphere data in any form is great. I would really like to have all the variables to further my knowledge as well.
Thanks for the contribution for sure.
nevergoodenuf
Well-Known Member
I don't know about sphere testing yet, but in a week or two, I will be getting the CLU058 3618 thermal tested on a couple of Starlites heat sinks. I just need to order two HLG 320s to test over 300 watts.
wietefras
Well-Known Member
Cool that you got all this data.
I guess the data actually quite closely matches the datasheet. It's just hard to see where on the datasheet to look since several parameters are not clear.
For instance, the data doesn't mention the Tc or Tj which would actually be quite important. It only mentions TMP(Deg) which is 150 (assuming Fahrenheit so 65C). That would probably be the ambient temperature in the sphere and then the COBs inside the fixture would be running much hotter. I've seen these pin coolers run at 50C above ambient. So that could mean a Tc of 115C. That would put the CXB3590 CD bin @1.4A on an efficacy of around 136lm/W according to the datasheet.
Although it's probably a combination of high Tc (say 85C) and the driver efficiency of 93%. That combination also gives you 136lm/W for the system according to the datasheet for the CD bin.
There are also these measurements which put the Gavita around that 1.72 mark as well (both also include driver losses):
In response to the Cree HortiCulti reference design, Gavita stated that putting a fixture inside a sphere is a bad idea. Makes sense since then much more light could get absorbed by the fixture. Flat plane integration would be better. Or at least mount it on the outside of the sphere shining in.
On the other hand, Cree also had the fixture inside their sphere and they arrived at 1.72PPF/W
I guess the data actually quite closely matches the datasheet. It's just hard to see where on the datasheet to look since several parameters are not clear.
For instance, the data doesn't mention the Tc or Tj which would actually be quite important. It only mentions TMP(Deg) which is 150 (assuming Fahrenheit so 65C). That would probably be the ambient temperature in the sphere and then the COBs inside the fixture would be running much hotter. I've seen these pin coolers run at 50C above ambient. So that could mean a Tc of 115C. That would put the CXB3590 CD bin @1.4A on an efficacy of around 136lm/W according to the datasheet.
Although it's probably a combination of high Tc (say 85C) and the driver efficiency of 93%. That combination also gives you 136lm/W for the system according to the datasheet for the CD bin.
I find it so odd that sphere data for this fixture can vary so wildly. Cree put a 1000W gavita Pro in their sphere (for the Horticulture Reference Design) and arrived at 1.72 PPF/W
There are also these measurements which put the Gavita around that 1.72 mark as well (both also include driver losses):
In response to the Cree HortiCulti reference design, Gavita stated that putting a fixture inside a sphere is a bad idea. Makes sense since then much more light could get absorbed by the fixture. Flat plane integration would be better. Or at least mount it on the outside of the sphere shining in.
On the other hand, Cree also had the fixture inside their sphere and they arrived at 1.72PPF/W
Actually, stardustsailor took thirty Bridgelux Vero 29's to a sphere owned by a University and the measurements matched the datasheets very closely. I think someone actually also put a few Cree COBs on a sphere, but I can't find that back so I could be wrong.
Last edited:
PSUAGRO.
Well-Known Member
Bugbee(Utah state) & beta test team(icmag) both got 1.72 & 1.75 ppf/w on the 1000w de, Nick told me his was "barely" used before being put in the sphere........don't know how bad the oxidation on the reflector was(OH is not near salt water,lol)? makes a big difference..........of course Gavita disagrees on those tests
his ul supply/calibrated for worst case scenario.........whatever the fuck that means, I need to get more info, he's probably sick of me by now! will see
Don't remember that with SDS having vero's tested at a Uni......I could be wrong as well
Last edited:
wietefras
Well-Known Member
@PSUAGRO, Yeah that's another point against reflectors. Still, 1.5 vs 1.7 seems a big difference just for possibly a bit of dirt on the reflector.
Gavita disagrees with that test because it doesn't match what they measure in actual applications. They have software which calculates light distribution based on number of fixtures and placement of the fixtures. That matches what they actually measure after the lights are installed. So they would know how much light is actually arriving at the plants and how much is coming from the fixtures.
Anyway, it was more an observation that you'd expect spheres to give definitive results but that apparently even those measurements can differ by 13%. Plus that perhaps methodology could be responsible for some of the difference.
Hmm can't find the SDS post with sphere results, but here is one where he refers to doing the sphere test:
http://rollitup.org/t/optic-lighting-vero-cobs-updated.860452/page-4#post-11330529
Gavita disagrees with that test because it doesn't match what they measure in actual applications. They have software which calculates light distribution based on number of fixtures and placement of the fixtures. That matches what they actually measure after the lights are installed. So they would know how much light is actually arriving at the plants and how much is coming from the fixtures.
Anyway, it was more an observation that you'd expect spheres to give definitive results but that apparently even those measurements can differ by 13%. Plus that perhaps methodology could be responsible for some of the difference.
Hmm can't find the SDS post with sphere results, but here is one where he refers to doing the sphere test:
http://rollitup.org/t/optic-lighting-vero-cobs-updated.860452/page-4#post-11330529
PSUAGRO.
Well-Known Member
OMG , never saw that!.............your right, thanks for pointing that out.............SDS was ahead of his time and we all still miss him.
I remember beta test team and whazzup(gavita) going at each other about that sphere test=== never came to an agreed upon conclusion about what is more accurate for their fixtures.......blaaahhhhh, so many variables.
edit.............I did see those SDS posts,lol
Last edited:
ya bongo
Well-Known Member
my point of view about top bins.
the CIE coordinates from the sphere for the two cxb3590
drawing this numbers into the Cree CIE-plot
and we clearly see the Light with reflectors is outside the Cree specs. -> blue line
the CIE coordinates from the sphere for the two cxb3590
drawing this numbers into the Cree CIE-plot
and we clearly see the Light with reflectors is outside the Cree specs. -> blue line
- Top Bin?
- color temp. changing with reflectors?
- a dirty cob surface that mostly filters blue light?
- a Cree fake?
sixstring2112
Well-Known Member
pretty cool to read through that thread,thanks
PSUAGRO.
Well-Known Member
It is a very good read...........man I have a shitty memory, even liked those SDS sphere posts back then,lol
I also forgot how sds and zang noticed increased output on the cxa's after some use.........very interesting
Last edited:
PSUAGRO.
Well-Known Member
he pops in every once in awhile.......... living the clean life now , no cigs, alcohol, sugar,salt,meat===planning on wife,house,family? last I heard
hope he's doing well in Hellas...