Top bin COB comparison

bassman999 · Dec 22, 2015

I am already upgrading from my CXA3070 ABs, well slowly
I am wondering which graph is considered closest to their efficiency?

Up to 5% difference in efficiency with the graphs at hgher currents

SupraSPL · Dec 23, 2015

The sheets I posted are based on "typical" figures at Tj 50C (junction temp). They take vF drop at 50C into account, temp droop and current droop. At higher current Tj 50C may not be realistic and at low current you can probably get below Tj 50C.

The other graph is based on Tc (case temp) but other than that I can't speak on how exactly the numbers were derived. According to the Cree sheet, at 1900mA, Tc 55C = Tj 85C.

Another way to put it, it is kind of a mess and once again we are left guessing to some extent. The same goes for the Vero PDFs if I recall, there is some confusion regarding which specs are based on junction temp and which are based on case temp.

SupraSPL · Dec 23, 2015

Mary's Confidant said:
I may have missed this but how do you determine what is "top bin" for various lights and spectrums?

I'm particularly interested in CXB 3590 3500s, 3070 3500s, 3070 4000s and 3070 5000s. I intend to do a mixture of spectrums using top bins for each but I don't know how we figure out CXB 3070 3500 top bin is BB? I know this is extremely rudimentary, so apologies to those who see this as "old hat".

The "top bin" can change as production methods improve. Initially the CXA3070 3000K top bin was Z2, but was replaced with the Z4 and eventually the AB. When the CXB3070 was release the 3000K top bin was now AD and it drew less power on top of that. So the Z2 was once considered top bin and could now be called low bin.

Also worth mentioning, there are nearly top bins and very top bins. For example @REALSTYLES got his hands on 7 of the CXB3590 6500K DD bin, which was the very top bin for 6500K and at the time very rare. The nearly top bin was the DB and you can now get it in color temp as low as 4000K 70 CRi from Cutter. The very top bins probably go to VIP customers that get first dibs and they may only show up in small quantities as left overs or samples. The other possibility is that they just mix in the super rare performers and they get binned lower than they really perform.

To answer your question, it makes sense that the CXB3070 3500K BB is the equivalent bin as the CXB3590 3500K CD in terms of the blue dies that are used. Because of it larger size and larger number of dies, all else being equal I would expect the 3590 CD to be slightly more efficient than the 3070 BB, and the curves and temp testing seems to confirm that. That said, @Greengenes707 had direct contact with CREE and was told otherwise, so that may be up for debate.

bassman999 · Dec 23, 2015

SupraSPL said:
The sheets I posted are based on "typical" figures at Tj 50C (junction temp). They take vF drop at 50C into account, temp droop and current droop. At higher current Tj 50C may not be realistic and at low current you can probably get below Tj 50C.

The other graph is based on Tc (case temp) but other than that I can't speak on how exactly the numbers were derived. According to the Cree sheet, at 1900mA, Tc 55C = Tj 85C.

Another way to put it, it is kind of a mess and once again we are left guessing to some extent. The same goes for the Vero PDFs if I recall, there is some confusion regarding which specs are based on junction temp and which are based on case temp.

Thanks for the reply.
I am not sure how to tell what my temps are other than the hs area by and in back of cob.
Not even sure where the other graph is from but was sad when I saw its numbers.

Stephenj37826 · Dec 23, 2015

SupraSPL said:
The "top bin" can change as production methods improve. Initially the CXA3070 3000K top bin was Z2, but was replaced with the Z4 and eventually the AB. When the CXB3070 was release the 3000K top bin was now AD and it drew less power on top of that. So the Z2 was once considered top bin and could now be called low bin.

Also worth mentioning, there are nearly top bins and very top bins. For example @REALSTYLES got his hands on 7 of the CXB3590 6500K DD bin, which was the very top bin for 6500K and at the time very rare. The nearly top bin was the DB and you can now get it in color temp as low as 4000K 70 CRi from Cutter. The very top bins probably go to VIP customers that get first dibs and they may only show up in small quantities as left overs or samples. The other possibility is that they just mix in the super rare performers and they get binned lower than they really perform.

View attachment 3570443

To answer your question, it makes sense that the CXB3070 3500K BB is the equivalent bin as the CXB3590 3500K CD in terms of the blue dies that are used. Because of it larger size and larger number of dies, all else being equal I would expect the 3590 CD to be slightly more efficient than the 3070 BB, and the curves and temp testing seems to confirm that. That said, @Greengenes707 had direct contact with CREE and was told otherwise, so that may be up for debate.

Hmm maybe if so that would definitely change things for me is cxb3070 bb was just as efficient as cxb3590 CD at same current/wattage.

Mary's Confidant · Dec 23, 2015

SupraSPL said:
The "top bin" can change as production methods improve. Initially the CXA3070 3000K top bin was Z2, but was replaced with the Z4 and eventually the AB. When the CXB3070 was release the 3000K top bin was now AD and it drew less power on top of that. So the Z2 was once considered top bin and could now be called low bin.

Also worth mentioning, there are nearly top bins and very top bins. For example @REALSTYLES got his hands on 7 of the CXB3590 6500K DD bin, which was the very top bin for 6500K and at the time very rare. The nearly top bin was the DB and you can now get it in color temp as low as 4000K 70 CRi from Cutter. The very top bins probably go to VIP customers that get first dibs and they may only show up in small quantities as left overs or samples. The other possibility is that they just mix in the super rare performers and they get binned lower than they really perform.

View attachment 3570443

To answer your question, it makes sense that the CXB3070 3500K BB is the equivalent bin as the CXB3590 3500K CD in terms of the blue dies that are used. Because of it larger size and larger number of dies, all else being equal I would expect the 3590 CD to be slightly more efficient than the 3070 BB, and the curves and temp testing seems to confirm that. That said, @Greengenes707 had direct contact with CREE and was told otherwise, so that may be up for debate.

Thanks for the detailed response Supra,

Is there any benefit to mixing spectra? I was going to purchase some top bin CXB3070 3500s, CXB3070 4000s and some CXB3590 5000s to mix together to create some type of synergy with the different temp profiles. But then I saw the video @Greengenes707 and @Growmau5 put together where they were saying the 3500 is nearly perfect.

Am I throwing away money on the 4k and 5ks that I could instead put into more 3500s (either 3070 or 3590)?

I'd really like to hear what you, @Greengenes707 and @Growmau5 (and others) have to say about this as I'd really like to pull the trigger soon but I don't want to spend my money foolishly trying to be "unique".

Are you guys supplementing with the smaller reds/violets anymore? Or does it appear unnecessary at this point?

wietefras · Dec 23, 2015

Sorry about the late response, but technically you could use a PAR meter for determining the PPF of a COB without an integrating sphere. You need to move the PAR meter in a quarter circle over the led measuring all the beam angles and then integrate the PPF from those measurements.

bassman999 · Dec 23, 2015

wietefras said:
Sorry about the late response, but technically you could use a PAR meter for determining the PPF of a COB without an integrating sphere. You need to move the PAR meter in a quarter circle over the led measuring all the beam angles and then integrate the PPF from those measurements.

I was wondering about doing it that way, but you would need to take numerous measurements.
How do you get results from that, just average them or use the mean?

Greengenes707 · Dec 23, 2015

wietefras said:
Sorry about the late response, but technically you could use a PAR meter for determining the PPF of a COB without an integrating sphere. You need to move the PAR meter in a quarter circle over the led measuring all the beam angles and then integrate the PPF from those measurements.

Not exactly. The closest thing is true flat plane integration. Which would take 100X more work to do than a sphere because the light is no longer confined...and you will have to measure IT ALL to be anywhere close to correct. And is not a totally scientifically recognized method.

Flat plane integration- Measurements were made in a dark room with flat black walls using a quantum sensor (LI-COR model LI-190, Lincoln, NE, USA), that was calibrated for each fixture with an NIST-traceable calibrated spectroradiometer. This calibration is necessary to correct for small spectral errors (± 3%) in the quantum sensor that occur because of imperfect matching of the ideal quantum response [16]. Measurements were made in three radial, straight lines below a level fixture and spatially integrated to determine total photon output. Measurements were made 2.5 cm apart near the center, increasing to 10 cm near the perimeter as PPF variation decreased. Fixtures were mounted 0.7 meters above the surface and measurements were made up to a 1.5 meter radius from the center and extrapolated farther using an exponential decay function. The flat-plane integration measurements were used to quantify the pattern of photon distribution from the fixture. Total fixture output from these measurements was similar to measurements made using an integrating sphere (Table 2). When redundant measurements were available, the integrating sphere measurements were used to quantify fixture efficiency.

** Notice the use of a decay function to extrapolate every photon all the way the true end of coverage(0umols).

***people please do not confuse this with the common "PAR map". FPI may be a map, but basic maps are not FPI.

wietefras · Dec 23, 2015

@bassman, You would need to integrate over the whole volume.

@Greengenes707, We are not talking about fixtures, but individual leds. Goniophotometers are an industry accepted standard for measuring light production from leds.

You don't need to measure it ALL either. Just one sweep over half of the beam angle (quarter circle). The light distribution from a COB is symmetrical.

Obviously it would be a challenge to build a goniophotometer at home and get any useful accuracy from an Apogee PAR meter, but technically it can be done.

You could build a simple one from lego

At least it could work to test PPF output differences between bins. That doesn't need to be so accurate in the absolute sense. Just needs to be repeatable to know which one is better. Although I guess a single PPFD measurement from each COB would already tell you that.

ketchup45685224 · Dec 23, 2015

Can anyone tell me if this any good?

http://growgreenmi.com/quantum-par-meter-micromols?gclid=Cj0KEQiAqemzBRDh2vGKmMnqoegBEiQAqJPuyADGtUWHMlryHlR9QuhHEd4pkMf_H_d__8ZpdMtgKvIaAv9d8P8HAQ

I'm trying to figure out just how much LED power is needed to get the same results as 43 watt/ft2 of a Hortilux 1000 watt HPS. I was planning on buying a PAR meter to measure the HPS and a 430 watt at the wall Vero 29 budget build. Then, compare them, do a little math, and start replacing my HPS. But, after reading this thread and doing a little bit of other research. I've learn all PAR Meter sensers aren't the same and most don't really measure the whole PAR spectrum. So, at best, these meters are only good for compare one white LED to another, not a HPS.

Everyone seems to be going on about grams per watt. I'm getting .8 - 1 gram/watt from my HPS of top shelf rock hard bud. I seen gardens running 600 watt HPS with higher grams per watt than mine, but the density isn't as good with the same cuts. Grams per is a good measure of efficiency but, efficiency comes in second to quality IMO. Can anyone tell me exactly how much LED power is needed or do I have to start experimenting?

nogod_ · Dec 23, 2015

~30w sqft of >50% efficiency leds

Many ways to achieve this....time to hit the books!

ketchup45685224 said:
Can anyone tell me if this any good?

http://growgreenmi.com/quantum-par-meter-micromols?gclid=Cj0KEQiAqemzBRDh2vGKmMnqoegBEiQAqJPuyADGtUWHMlryHlR9QuhHEd4pkMf_H_d__8ZpdMtgKvIaAv9d8P8HAQ

I'm trying to figure out just how much LED power is needed to get the same results as 43 watt/ft2 of a Hortilux 1000 watt HPS. I was planning on buying a PAR meter to measure the HPS and a 430 watt at the wall Vero 29 budget build. Then, compare them, do a little math, and start replacing my HPS. But, after reading this thread and doing a little bit of other research. I've learn all PAR Meter sensers aren't the same and most don't really measure the whole PAR spectrum. So, at best, these meters are only good for compare one white LED to another, not a HPS.

Everyone seems to be going on about grams per watt. I'm getting .8 - 1 gram/watt from my HPS of top shelf rock hard bud. I seen gardens running 600 watt HPS with higher grams per watt than mine, but the density isn't as good with the same cuts. Grams per is a good measure of efficiency but, efficiency comes in second to quality IMO. Can anyone tell me exactly how much LED power is needed or do I have to start experimenting?

bassman999 · Dec 23, 2015

Greengenes707 said:
Not exactly. The closest thing is true flat plane integration. Which would take 100X more work to do than a sphere because the light is no longer confined...and you will have to measure IT ALL to be anywhere close to correct. And is not a totally scientifically recognized method.

Flat plane integration- Measurements were made in a dark room with flat black walls using a quantum sensor (LI-COR model LI-190, Lincoln, NE, USA), that was calibrated for each fixture with an NIST-traceable calibrated spectroradiometer. This calibration is necessary to correct for small spectral errors (± 3%) in the quantum sensor that occur because of imperfect matching of the ideal quantum response [16]. Measurements were made in three radial, straight lines below a level fixture and spatially integrated to determine total photon output. Measurements were made 2.5 cm apart near the center, increasing to 10 cm near the perimeter as PPF variation decreased. Fixtures were mounted 0.7 meters above the surface and measurements were made up to a 1.5 meter radius from the center and extrapolated farther using an exponential decay function. The flat-plane integration measurements were used to quantify the pattern of photon distribution from the fixture. Total fixture output from these measurements was similar to measurements made using an integrating sphere (Table 2). When redundant measurements were available, the integrating sphere measurements were used to quantify fixture efficiency.

** Notice the use of a decay function to extrapolate every photon all the way the true end of coverage(0umols).

***people please do not confuse this with the common "PAR map". FPI may be a map, but basic maps are not FPI.

If it is too hard and/or expensive to measure everywhere how about doing comparative testing using optics? The same testing with the same optics on each cob, and there s no stray light to worry about.
The optics loss is then added back to the results...?
The problem I see is finding/or making universal optics/reflector

dandyrandy · Dec 23, 2015

bassman999 said:
My take is that I can run more 3070s and run softer than the 3590 and get close efficiency with better spread for a similar or better price of purchase

I just changed my plan of 16 cxb3590's @ ~

ketchup45685224 said:
Can anyone tell me if this any good?

http://growgreenmi.com/quantum-par-meter-micromols?gclid=Cj0KEQiAqemzBRDh2vGKmMnqoegBEiQAqJPuyADGtUWHMlryHlR9QuhHEd4pkMf_H_d__8ZpdMtgKvIaAv9d8P8HAQ

I'm trying to figure out just how much LED power is needed to get the same results as 43 watt/ft2 of a Hortilux 1000 watt HPS. I was planning on buying a PAR meter to measure the HPS and a 430 watt at the wall Vero 29 budget build. Then, compare them, do a little math, and start replacing my HPS. But, after reading this thread and doing a little bit of other research. I've learn all PAR Meter sensers aren't the same and most don't really measure the whole PAR spectrum. So, at best, these meters are only good for compare one white LED to another, not a HPS.

Everyone seems to be going on about grams per watt. I'm getting .8 - 1 gram/watt from my HPS of top shelf rock hard bud. I seen gardens running 600 watt HPS with higher grams per watt than mine, but the density isn't as good with the same cuts. Grams per is a good measure of efficiency but, efficiency comes in second to quality IMO. Can anyone tell me exactly how much LED power is needed or do I have to start experimenting?

I run 800w of Vero 29 cobs in about 11sqft. Donkey colas solid buds. The avatar is tda 3d and blue dream. And some SSH somewhere in there. I'm with you on quality. But I like to save a buck. I ordered 16 cxb3070 and will add 16 more this summer. 20w each?

Greengenes707 · Dec 23, 2015

I stand by the actual

bassman999 said:
If it is too hard and/or expensive to measure everywhere how about doing comparative testing using optics? The same testing with the same optics on each cob, and there s no stray light to worry about.
The optics loss is then added back to the results...?
The problem I see is finding/or making universal optics/reflector

Ya, totally fine and useful. That is a relative test and what most meters are perfect for. Precision will be there, and it's the precision is the important part in a relative test. It can be inaccurate by a lot...as long as it's consistent in it's in accuracy...that is/equals precision. Assuming no color shift with the optics(99%shouldn't), it's a very valid test.

Absolute values are where things have to really get measured correctly. I have measured single die, cobs and full fixtures...none emit a perfectly symmetrical pattern. It may be just a few µmols, but it's measurable and should be accounted for if looking for the almighty true facts....or any kind of tie breaker.

bassman999 · Dec 23, 2015

Greengenes707 said:
I stand by the actual

Ya, totally fine and useful. That is a relative test and what most meters are perfect for. Precision will be there, and it's the precision is the important part in a relative test. It can be inaccurate by a lot...as long as it's consistent in it's in accuracy...that is/equals precision. Assuming no color shift with the optics(99%shouldn't), it's a very valid test.

Absolute values are where things have to really get measured correctly. I have measured single die, cobs and full fixtures...none emit a perfectly symmetrical pattern. It may be just a few µmols, but it's measurable and should be accounted for if looking for the almighty true facts....or any kind of tie breaker.

All of this testing is above my level of experience.
All the work done by everyone here that is shared is very appreciated!

ketchup45685224 · Dec 23, 2015

can someone show me the math behind "30w sqft of >50% efficiency leds" I've googled it and found so far is it's like comparing apples to oranges as far as units are concerned...

nogod_ · Dec 23, 2015

Well, your HID is 30-35% efficient so if you are at 43w/sqft that means you are roughly 15 parwatt/sqft. 15 watts emitted as light, 28w radiated as heat.

30w of CoB @50% means you are at 15parwatts/sqft....15 watts emitted as light, 15w radiated as heat.

So you are emitting the same amount of light for less power draw.

Rahz · Dec 23, 2015

37 watts per foot at 25% efficiency = 37 x .25 = 9.25 Par watts

60 x .25 = 15 par watts, 45 watts of heat

ketchup45685224 · Dec 23, 2015

thanks, I suppose that is a good answer for the real world. I was asking how you guys know 50% of the power put into a LED comes out as PAR watts. But maybe that math would be getting kind of crazy and I really don't need to know that. :-)

Top bin COB comparison

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member