PICOGRAV
Well-Known Member
Why don't you use a more powerful series with reflectors?
PICO's DIY Thread - Advise, Ideas and Technology - NO PANEL REPS!
- Thread starter PICOGRAV
- Start date
Why don't you use a more powerful series with reflectors?
PICOGRAV
Well-Known Member
No, this is not ideal.
Hi picograv.
initially I did plan on running fewer larger chips such a 4 vero 29 but couldn't find reasonably priced drivers. I then started looking at the smaller vero chips and trying to work out the best positioning and amounts needed. I then looked at using 9 vero 18 in three rows of three and then 16 vero 13 or 10 in 4 rows of four and finally came to what you see above.
Nothing is set in stone until I place my order and that wont be until I know everything I need and exactly I want.
I will do a few calculations and work out what drivers and chips give the best lumen/watt/£ but still remain efficient.
Could I ask what you would recommend for which k you would recommend? mr flux posted a graph in my thread showing the vero 4000k against a cree xte 3700k and the vero looked very promising.
many thanks andy.
initially I did plan on running fewer larger chips such a 4 vero 29 but couldn't find reasonably priced drivers. I then started looking at the smaller vero chips and trying to work out the best positioning and amounts needed. I then looked at using 9 vero 18 in three rows of three and then 16 vero 13 or 10 in 4 rows of four and finally came to what you see above.
Nothing is set in stone until I place my order and that wont be until I know everything I need and exactly I want.
I will do a few calculations and work out what drivers and chips give the best lumen/watt/£ but still remain efficient.
Could I ask what you would recommend for which k you would recommend? mr flux posted a graph in my thread showing the vero 4000k against a cree xte 3700k and the vero looked very promising.
many thanks andy.
NewGrower2011
Well-Known Member
I went and checked the specs and the Vero's below the 29 it said 2.x the rated for those models... But wouldn't that be at progressively less efficiency or are they saying you can drive 2x the nominal and at the rated efficiency? I'm a software guy so some of this still alludes me currently... I dabbled in electronics as kid so I know some very basic stuff...
NewGrower2011
Well-Known Member
I'm thinking those Vero 18's or smaller with multiple chips is looking like something I'd try perhaps. I'm not sure if I'd go and supplement with additional blue or far-red channels... Still doing my homework...
What all should I factor in when deciding between say the 10/13/18... i.e. X number of 10's versus X number of 13's... If I'm looking to spread my footprint out, minimize costs (assuming higher density wafer == higher cost per unit) up front and by being more 'modular' if a given part/channel fails I don't lose the entire array... etc...
I need to go back and re-read but what all do I need to figure out driver wise when doing arrays of the Vero chips... i.e. figuring out series/parallel and what mA drivers and # of? Any good primers?
What all should I factor in when deciding between say the 10/13/18... i.e. X number of 10's versus X number of 13's... If I'm looking to spread my footprint out, minimize costs (assuming higher density wafer == higher cost per unit) up front and by being more 'modular' if a given part/channel fails I don't lose the entire array... etc...
I need to go back and re-read but what all do I need to figure out driver wise when doing arrays of the Vero chips... i.e. figuring out series/parallel and what mA drivers and # of? Any good primers?
PICOGRAV
Well-Known Member
http://www.bridgelux.com/wp-content/uploads/2013/10/DS33-Bridgelux-Vero-29-Datasheet-2013.07.18.pdf
The other veros can be run at 2x drive current, not the vero 29
PICOGRAV
Well-Known Member
I really feel the 4000K is the best option for the blue light, I would use the 5000K again if the CRI was at least 80, you want the control of the higher light temperature, especially if you are going to train your plants.
I would only use the Vero 18s of you are going with the super high CRI versions as they are not available in the vero 29 or if your building a side light, the smaller packages would be useful.
As for driving them, you have to think outside the box and don't get hung up on the "traditional" CC LED drivers out there, regulated power supplies with voltage and current control will always allow your self more freedom in the final design...
NewGrower2011
Well-Known Member
Are computer PSU's of any use in the DIY LED pursuits? I have a bunch of those... old server-grade ones even and some newer ones from higher-end gaming rigs... I'm thinking no but I'll ask all the same...
NewGrower2011
Well-Known Member
Right... I was saying the ones besides the 29 could be done @ 2.x - I understood that the 29 was capped at 1.5x.... Sorry if I wasn't clear...
NewGrower2011
Well-Known Member
Also out of due diligence and knowing what is out there... what is comparable to these from other big vendor camps? Anything even comparable? I've heard of terms like dense matrix array, COB, etc... So I'm guess there's something from Phillips, Cree, etc...
PICOGRAV
Well-Known Member
Veros seem to be the best offering right now...
smokey the cat
Well-Known Member
Depends what you wanna do, right? More output or higher efficiency. We gotta accept that driving at race speeds we're not gonna get the same fuel efficiency out of the wagon. Can't imagine you spend your nights lying awake worrying that because you run them harder than nominal you might not get the full 50,000 hours out of your emitter, lmao.
Gotta admire the man who drives cars fast. But doesn't stop us sperging about fuel efficiency
---
Question for Pico as you're the guy with the most direct Vero experience in here. How do you find the emitter heat output as you increase current? I know you've got an awesome server wind tunnel setup, but what are your thoughts of these things as heat producers?
Oh - and how flat was the base on yours?
Thanks mate
picograv What would you use to control the dimming function if not the 100komh pot.
If you look at the last page of the the data sheet it states a 100k resistor
http://datasheet.octopart.com/LPF-60D-36-Mean-Well-datasheet-12548763.pdf
If you look at the last page of the the data sheet it states a 100k resistor
http://datasheet.octopart.com/LPF-60D-36-Mean-Well-datasheet-12548763.pdf
SupraSPL
Well-Known Member
You can run them as hard as you want. If you run them at full power the efficiency is so low you'd be better off with HPS and as SMokey alluded to, your emitter will lose its output over time despite drawing the same amount of power. Those are the facts based on the Bridgelux PDF.
smokey the cat
Well-Known Member
Supra, at what point do you think the crossover to HPS-level efficiency is? You think a straight comparison of quoted lumens to lumens gives an accurate comparison for between Led/HPS for our use?
At 'max'-3150mA the V-29 low-CRI 3000k shows 110lm/w, 'nominal'-2150mA gives 120lm/w, 700mA gives 144lm/w.
It is definitely fair bit more efficient when soft.
At 'max'-3150mA the V-29 low-CRI 3000k shows 110lm/w, 'nominal'-2150mA gives 120lm/w, 700mA gives 144lm/w.
It is definitely fair bit more efficient when soft.
Positivity
Well-Known Member
A 600w hps put out about 90,000 lumens. So at 700ma it would take 625w of vero 29 to hit that many lumens. I'd like to see that in action..
SupraSPL
Well-Known Member
Assuming you aggressively cool them and taking into account junction temp losses and current droop, here are the numbers I am going by:
Vero 29 3000K 80CRI
.7A 127 lm/w 37.6%
1.4A 115 lm/w 34%
2.1A 105 lm/W 31.25%
2.8A 93.45 lm/w 27.8%
3.15A 89 lm/w 26.4%
Those lumen numbers do give some guidance especially when comparing against other 3000K LEDs (HPS is about 2000K), but I prefer to look at things in terms of radiometric efficiency. A brand new 600 HPS bulb is about 40% efficient and that why it has been so hard to beat. The main LED advantage is that it shoots all those photons in a cone pattern as opposed to a globe pattern and if you run them at reasonable junction temps they will maintain 90-95% of their efficiency for 10 years. We can now build LED lamps that have an overall efficiency of 45% at 700mA but they are pricey and time consuming to build.
Vero 29 3000K 80CRI
.7A 127 lm/w 37.6%
1.4A 115 lm/w 34%
2.1A 105 lm/W 31.25%
2.8A 93.45 lm/w 27.8%
3.15A 89 lm/w 26.4%
Those lumen numbers do give some guidance especially when comparing against other 3000K LEDs (HPS is about 2000K), but I prefer to look at things in terms of radiometric efficiency. A brand new 600 HPS bulb is about 40% efficient and that why it has been so hard to beat. The main LED advantage is that it shoots all those photons in a cone pattern as opposed to a globe pattern and if you run them at reasonable junction temps they will maintain 90-95% of their efficiency for 10 years. We can now build LED lamps that have an overall efficiency of 45% at 700mA but they are pricey and time consuming to build.
PetFlora
Well-Known Member
Keep in mind that any globe that is placed horizontally does not provide 100% directly to the canopy + the radiant heat means bulb must be kept a lot farther from the canopy, so... a quality 300-350w led panel should be ~ equal o a 600w hps bulb. And as a bonus, no hood or exhaust fan is needed
smokey the cat
Well-Known Member
My two Vero arrived today from Newark - but there was a problem. Everything was boxed properly, but one of the emitters was just chucked between two foam sheets - no antistatic packaging at all.
I guess I can let you all know how sensitive to static gremlins these things are.
As I'm using a freight forwarder I'm actually paying more for postage than for the Vero themselves. If it's bad I'll just have to lump it and order another. Still cheaper then getting them locally - the local distributor adds 600-700% markup ontop of US pricing.
My question for experienced LED gurus: is there a fast easy method to test these?
Wondering if I can I chuck one or two 9V battery across the terminals or something - thinking 9-18V it might be enough juice to get a glow from a 30V part. Else I'll need to wait till I can get these AC drivers half sorted and organise some sort of heatsink to test with as my ebay one isn't here yet.
I just don't wanna release the magic smoke! http://en.wikipedia.org/wiki/Magic_smoke
I guess I can let you all know how sensitive to static gremlins these things are.
As I'm using a freight forwarder I'm actually paying more for postage than for the Vero themselves. If it's bad I'll just have to lump it and order another. Still cheaper then getting them locally - the local distributor adds 600-700% markup ontop of US pricing.
My question for experienced LED gurus: is there a fast easy method to test these?
Wondering if I can I chuck one or two 9V battery across the terminals or something - thinking 9-18V it might be enough juice to get a glow from a 30V part. Else I'll need to wait till I can get these AC drivers half sorted and organise some sort of heatsink to test with as my ebay one isn't here yet.
I just don't wanna release the magic smoke! http://en.wikipedia.org/wiki/Magic_smoke
Someacdude
Active Member
So this isnt the part where i ask what kind of leds i need right?
