Can get more addictive ..Can it ?

Good info SDS. To back that up, I noted a decrease of only 3.2% light when running the Vero29 3000K at 2.23A on a Rosewill RCX-Z1 CPU cooler. With a slight active cooling (1.3W), the hottest part of the heatsink I could measure was 34C. The Vf change from 23C cold start to thermally stable was .56V.

At .016V/C, that should translate into a Tj of ~58C. Awesome

So ...
In the Vero datasheet ....


0.946 of 25C 'Typical' Flux .

'Typical' Flux at 25 C is 9900 lm .
Flux is measured from a pulsed Vero 29 ,in order to keep Tc =25 C .
(Not to heat up ) .
So 0.946 * 9900 = 9365.4 lm

Vero 29 has a typical Vf 0f -16 mV / C ..

so ...
55-25 = 30 C
30 * 0.016 =0.48 V
From typical 25C Vf of 38 VDC ,subtracting 0.48 VDC => Vf= 37.52 V ...
LER= 321
Io=2100 mA
Tc=55 C

so ....


37% radiant efficiency ...
At 2100 mA and with aTc=55 C .

Bye-bye HPS ...
No matter the type or wattage ..
The 600W of HPSs have highest efficiency figures ,close to 33-35% ,
only when brand new and for a short while after,~1500-2000 hours .
For the 400 & 250 watt HPS ,better assume a rad.efficiency of ~30%..At best case ...
And this dual COB fixture is a direct high-tech substitute or 'competitor' of the latter HPS power class ...

( Most of growers will be getting about 200 gr from every grow cycle with a 400 W HPS .
Same most growers will be getting about the same yield ,with half the wattage ,using two Vero29 at 2100 mA ..Around 0.8 - 1gpw ..Maybe a 'better' spectum accounts for that constant difference between the two techs used ..Light power distribution is pretty similar at both cases,I would hardly guess ...Nothing else is left to be accounted for ..Radiated heat from HPS ?
It's bull that forces plants to assimilate more nutes ..Bull ..They take up more water...
That's why the HPS buds lose much weight / size when drying - 75% w/w ,while the SSL COB buds lose pretty much less weight / size when drying ...55-65 % w/w
They 'hold' less water-load ,first -place ...)

At a Ta =35 C (...hot summer days ) ,junction temperature will be around 70-75 C ..
Even then ,the service life of the fixture ,it extends more than 50.000 hours .

4x7x18=504
10x7x12=840
Average grow cycle = 1344 hours .Say 1400 h.

Minimum number of full grow cycles ,per COB arrays' service life (LM 80 ): 36 .

For every 3-6 grow cycles one new HPS is needed ..

for 36 grow cycles a min. of 6x HPS lamps will be needed .
6 x $50 = $300 ..
Thats about half of the (...'retail' let us just say ) price of the COB fixture ...\
Add a crappy ballast of $50 ,a $10 reflector ,wiring and electricity cost difference for 36x grows (that's gonna be plenty ),
to the HID option ...
And that is your best-best -best ,as cheapest of cheapest -@$$€$ , case / option with HPS..

Spending ~$400 one can DIY make a dual-COB fixture exactly like that ...
(I need kinda more ,in order to make it for him/her ,instead ... ..But not much ..)


Cheers.

Love it guys. I think that your fixture is very traditional in the way it's used...but with high tech engines doing the work.
Just for clarification an hps LER is 380 +\- . 600hps is hitting 158lm/w or 41.5% effciency.

Greengenes707 said:

Love it guys. I think that your fixture is very traditional in the way it's used...but with high tech engines doing the work.
Just for clarification an hps LER is 380 +\- . 600hps is hitting 158lm/w or 41.5% effciency.

Click to expand...

380+ LER means that most of power is concentrated near 555 nm green ..
Even with 41.5 % rad efficiency ...
COBs of 3K -80Ra have so much better spectral power distribution ,
even from a blue -wl -band enhanced HPS like SON-T Agro or GreenPower.
Anyway ...
It's been so long from last time I dealt with HPS ...
(No ,this is not a good excuse for being mistaken ,I know .. )

At least for the 400 W class ,I'm right ...
Radiant Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 W
Radiant Efficiency. . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 MW/W
Conversion Factor . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 MW/Im

http://www.usa.lighting.philips.com/connect/tools_literature/downloads/p-2707b.pdf
http://buymarijuanaseeds.com/community/threads/bulbs-comparison-tool.89513/

Yes ,this fixture is pretty 'trad' ,in its use ...
As I've already mentioned ...A direct replacement for 250-400 Watt HPS .

That is something I have been wondering about. @Richelsdorfite digitized and analyzed the HPS graph for us and came up with this:

Richelsdorfite said:

For Standard HPS - 2200K - 140000 lumens - 1000W, I find :
- Radiant Power 399.79 Watt & 2006.62 umol/s - LER = 350

Click to expand...

SupraSPL said:

Thanks RD! Is that for PAR watts 400nm-700nm?

Click to expand...

Richelsdorfite said:

For PAR 400-700nm the radiant flux is : 358 Watts & 1750 umol/s

Click to expand...

So in the visible range, it is 35.8% efficient, and apparently 4.2% is out of that range? Now they are claiming 155,000 lumens but they may have adjusted the spectrum.

...thanks supra... ...not 400 ..not 600.. its for 1000w hps... wow...

...whats the best 600 or 1000 hps?...my curiosity its only for saw her graphs...

pd... and thanks R for digitized and analyzed that data...

SupraSPL said:

That is something I have been wondering about. @Richelsdorfite digitized and analyzed the HPS graph for us and came up with this:







So in the visible range, it is 35.8% efficient, and apparently 4.2% is out of that range? Now they are claiming 155,000 lumens but they may have adjusted the spectrum.

Click to expand...

In originally had a 388ler...worked backwards from 36% effciency. Then found 380LER stated online somewhere twice.
From his numbers the LER increased when confined to the visible. With so much going on for hps past 700nm...it's hard to use lm/w as a variable for the raw efficiency(all nm's)
I am assuming the huge IR spike and little uv contribute the near 40% raw(all nm's) effciency vs the visible. And has no luminous function values but is a significant amount of watts though.

SDS, I have some interesting Vero29 vs CXA3070 data. All of these measurements were pulsed at 23C so temp dropp would not interfere. I monitored the lux, current and Vf during the pulses and have a high degree of confidence in all the measurements.

Vero29@ 2.24A = 87.27W = 2850 lux
Vero29 @ .151A = 5.13W = 224 lux

CXA3070 @ 2.24A = 88.97W = 3000 lux
CXA3070 @ .152A = 4.85W = 248 lux

What I am looking at here is the current droop.
The Vero was dissipating 5.88% power and sustained 7.86% of its light. (1.34 ratio)
The CXA was dissipating 5.45% power and sustained 8.27% of its light. (1.52 ratio)

I would have expected the CXA to outperform the Vero a lot more when it comes to current droop?
According to the Cree PCT, at 13.4% power the CXA3070 should sustain 20.4% of its light (350mA vs 2.2A), a ratio of 1.52. Maybe current droop is not linear, especially at lower currents?

SupraSPL said:

SDS, I have some interesting Vero29 vs CXA3070 data. All of these measurements were pulsed at 23C so temp dropp would not interfere. I monitored the lux, current and Vf during the pulses and have a high degree of confidence in all the measurements.

Vero29@ 2.24A = 87.27W = 2850 lux
Vero29 @ .151A = 5.13W = 224 lux

CXA3070 @ 2.24A = 88.97W = 3000 lux
CXA3070 @ .152A = 4.85W = 248 lux

What I am looking at here is the current droop.
The Vero was dissipating 5.88% power and sustained 7.86% of its light.
The CXA was dissipating 5.45% power and sustained 8.27% of its light.

I would have expected the CXA to outperform the Vero a lot more when it comes to current droop?

Click to expand...

And I've some data about Vero29 At 1400 mA and Tc=40C

..

Both arrays seem to be almost identical ,regarding their efficiency and spectrum,at least .

Cheers.

Not to be misunderstood, I am not arguing about which is better, just exploring the characteristics of each architecture and trying to find how to best use each to its strengths.

Those lux measurements are not meant to compare output. I recognize that the Vero has more red output and that will eliminate the ability of the lux meter to make a comparison between the CXA and the Vero. I am just focusing on the current droop in this case.

SupraSPL said:

Not to be misunderstood, I am not arguing about which is better, just exploring the characteristics of each architecture and trying to find how to best use each to its strengths.

Those lux measurements are not meant to compare output. I recognize that the Vero has more red output and that will eliminate the ability of the lux meter to make a comparison between the CXA and the Vero. I am just focusing on the current droop in this case.

Click to expand...

Vero has a more 'linear ' current droop .
(Both at Tc=25 C ).
It seems that CXA3070 declines a tad more in performance than Vero29, as If increases .
While at low currents,the CXA has a 'steeper' performance increase towards 100% .
So when driven lower than 100% the CXA3070 shows a significantly less 'current droop ',
than Vero29.
But when driven higher than 100% ,then CXA3070 has a more steep performance decrease than Vero29,with the latter having a more 'linear' current droop ,thus a more 'stable droop behaviour' ,than CXA3070.
Clearly the CXA3070 ,performs at it's best ,when driven lower than ~1800 mA .
From the other hand ,Vero29 ,seems not to have any particular problems with abnormal 'droop behaviour ' ,almost throughout all of it's driving range.

And their thermal behaviour compared .

Vero29


CXA3070 (@1400 mA )



Combined ...


Clearly Vero29 is better than the CXA3070 ,when it comes in maintaining high output versus Tc increasing.
Vero has a more thermally conductive substrate material (aluminium ) than the CXA has (ceramic) .
Also it has larger heat transfer surface area,than the CXA3070 has ... ~263 mm^2 more ,actually or ~ 35% more heat conduction surface area .

Vero also has 24 more blue excitation diodes than the CXA3070 (156 vs 132 ) ,
but more spaced apart inside the package and under a larger LES.
Larger LES also ,helps keeping phosphor particles cooler ,thus providing more stable CCT over time and having
less phosphor degradation .

As soon CXAs case hits 45C ,output is already at 95% of nominal .
Vero needs a Tc of 55C ,to output the same 95% of nominal .
A whopping 10 C difference .

Vero29 is much more 'stable ' regarding output % vs Tc ,than the CXA3070.
And 'easier' to 'maintain' cool and efficient enough ,of course .

Thank you SDS, that is exactly what I needed to see to finalize my decision in going with the Vero 29's......

Only thing that could persuade me to get CXA3070 is if I could find anything other then a Z2 Bin.

Great info thanks for all the hard work on that SDS!

Reefer, if you are going to run hard you would be better off with the Vero29, even if you could get the AB bin. But if you are really shooting for max efficiency and you want to run soft, the CXA3070 AB is back in stock for $42 shipped. This is the same seller that I got a genuine AB bin sample from. I am guessing these will sell out fast so make sure you specify that you want the 3000K AB when you order in case they are already sold out.

http://www.aliexpress.com/item/Freeshipping-1pcs-Cree-XLamp-CXA3070-led-74-117W-COB-EasyWhite-5000K-Warm-White-3000K-LED-Chip/2052849889.html

Supra you dog.

Now I might have to get 2 CXA3070 AB bins, and 2 of the Vero 29's just for shiggles

My plan is too run them at 1.5 A +/- 5% I think that gives me pretty good dissipation for a 2x2 tent.

If I don't feel like I am getting what I want the Vero's will be pumped up to 2.1 A. But I think the CXA's will stay at the 1.5 A purely for efficiency.

So,by now " V series" units will be manually power adjusted ,
by a 12 position rotary switch ,instead of a pot .

Analytical ,every ( preset ) switch position of the unit .
(TOT. = 2x Vero29 , cells with diagonal lines are just for calculation purposes )

** V series units ,utilise 2x Bridgelux BXRC-30E10K0-L-03 chip-on-board L.E.D. arrays.
( Distribution of photosynthetic quantum flux :10,6% @400-499 nm || 40,7% @ 500-599 nm || 44% @600-699 nm || 4.7% @700-800 nm .Phytochrome Action : R/FR= 6,27 || PPE= 65,72% )

Unit size :320 x 255 x 130 mm (W x D x H )
A single unit replaces a 250 W HPS .
Dual -unit set up replaces a 400 W HPS .
And so on ...

Cheers.

And here's some efficiency figures (light source / system ) of some other light systems ...













Sylvania

My first Vero29 measured 38.5Vf warmed up at 2.24A, higher than I expected based on the Vf graph. Not sure if that is a fluke because I have only tested one so far. Do yours run a but high in Vf also?

POWER ADJUST INTERFACE -Revised

(Overall system efficiency included => Radiant power / Plug power ) :


It seems that in switch positions 4-5-6 ( Io= 800-1150 mA ),a V-series unit operates at it's best ,efficiency-wise...
(similar to 600W HPS system efficiencies )
While at positions 7-8-9 ( Io= 1300-1700 mA ),overall system efficiency
is similar to the best 400 W HPS system efficiencies .

Hm...Good to know ! ..
Cheers.

SupraSPL said:

My first Vero29 measured 38.5Vf warmed up at 2.24A, higher than I expected based on the Vf graph. Not sure if that is a fluke because I have only tested one so far. Do yours run a but high in Vf also?

Click to expand...

In what specific moment have you tested the Vf ?
After the system reached it's thermal equilibrium ?
..Anyway ..
Don't forget that once, there was a Vf binning regarding leds ,also ...
Veros are binned only for the McAdam steps ..For their chromaticity ..
Not any power bins ,not any Vf bins ....

Cheers.