reefgeezer
Active member
Coming up on upgrade time. My current lights are 3W Crees in series with a Meanwell driver & manual dimmers. So... what is state of the art for DIY LED fixtures these days?
LED Trends?
- Thread starter reefgeezer
- Start date
hi,
arduino or something like that for pwm control and ldd's of course. i dont know if you know it, this my diy arduino thing:
http://www.youtube.com/watch?v=P7i0LkUfnf4
arduino is rocksolid stable and makes a lot of fun to code..:bigeyes:
reefgeezer
Active member
So... right now I have 5 sets of 12 X 3W Crees wired in series to a Meanwell Drive (-ELN I think). 2 sets are RB - 1 is CW - 1 is 4CW, 2WW and 6B - and 1 is 8UV, 2R, & 2G. They are arranged in parallel lines rather than clusters. I don't run the strings at anywhere near max current and I can manually dim each of the strings and do so to get the color rendition I like. The white strings are usually dimmed almost off.
If the LEDs I have are still decent choices - I suppose the upgarde could be getting drivers that will allow PWM control and learning what the heck arduino's are and how to code! I'm old though, it might take a while.
If the LEDs I have are still decent choices - I suppose the upgarde could be getting drivers that will allow PWM control and learning what the heck arduino's are and how to code! I'm old though, it might take a while.
reefgeezer
Active member
So cool! I'd like to do some of that. I guess I better start reading up on this stuff.
i may be wrong because i'm not an expert on electronics, but i think you can instead of connecting the strings directly to the psu you connect them to a ldd driver for each channel. the ldd driver has a pwm input that you can connect directly to an arduino.
about the arduino: dont be afraid, it is not too hard
the possibilities you have with it are endless.reefgeezer
Active member
i may be wrong because i'm not an expert on electronics, but i think you can instead of connecting the strings directly to the psu you connect them to a ldd driver for each channel. the ldd driver has a pwm input that you can connect directly to an arduino.
about the arduino: dont be afraid, it is not too hard
Great! Sounds like the right direction. Time to plan & study. Thanks everyone. One last question... Is RapidLED.com still the place to buy LED related stuff?
and they are 150 for two pucksUtah Fatman
New member
Absolutely outstanding. Nice job.
ReefLedLab
In Memoriam
Agreed. I would like to add - the worst royal blue bin of Luxeon T has a better efficacy than best bin of any Cree LEDs. This is will surprisingly for many reefkeepers, isn't it?
Not wanting to be picky, but philips quote typical of 53% and cree quote up to 53% (for the XTE), both at 85° - OK that still reflects better on the luxeons than the XTEs, but only a little, not as bad as your quote implies. And over here in the UK, I can get decent bin XTEs mounted for very little. Up to yet, the same cannot be said for luxeon Ts
Tim
ReefLedLab
In Memoriam
Hi, Tim.
Your opinion is widespread. But let's check it with datasheet. For example for Royal Blue. From Cree side will be XT-E, from Philips side will be Luxeon T.
The best bin of luminous flux E of Luxeon T is extremely rare, so we will not calculate for it. But not only bin E of Luxeon T is outperform Cree RB, but also bin D, it available on open market without problem. But let's look on bin C, it is more interesting, because it widespread and very easy to buy. Luxeon T RB bin C has a approx. 1075mWt on 700mA and forward voltage (FW) 2.8V. So we can see efficacy 1075/(2.8*700)*100=55%. Let's look at the best bin of Cree. Best RB Cree XT-E can output 600mWt at current 350mA and approx. 1050mWt at 700mA. FW will be 3.2V, so we have efficacy 1.050/(0.7*3.2)*100=47%. Let's check efficacy of bin B of Luxeon T: 1025/(2.8*700)*100=52%! Let's look at the worst bin of Lux T: 975/(2.8*700)*100=50%!
We can make simple conclusion - the WORST bin of Luxeon is noticeable better than BEST bin of Cree.
Your opinion is widespread. But let's check it with datasheet. For example for Royal Blue. From Cree side will be XT-E, from Philips side will be Luxeon T.
The best bin of luminous flux E of Luxeon T is extremely rare, so we will not calculate for it. But not only bin E of Luxeon T is outperform Cree RB, but also bin D, it available on open market without problem. But let's look on bin C, it is more interesting, because it widespread and very easy to buy. Luxeon T RB bin C has a approx. 1075mWt on 700mA and forward voltage (FW) 2.8V. So we can see efficacy 1075/(2.8*700)*100=55%. Let's look at the best bin of Cree. Best RB Cree XT-E can output 600mWt at current 350mA and approx. 1050mWt at 700mA. FW will be 3.2V, so we have efficacy 1.050/(0.7*3.2)*100=47%. Let's check efficacy of bin B of Luxeon T: 1025/(2.8*700)*100=52%! Let's look at the worst bin of Lux T: 975/(2.8*700)*100=50%!
We can make simple conclusion - the WORST bin of Luxeon is noticeable better than BEST bin of Cree.
Not quite sure what you mean by my "opinion is widespread"
The figures I quoted were from the datasheets for the luxeon T from the Philips web site and the Cree web site for the XT-E. And, as I say, the luxeon Ts are anything but easy to get, over here in the UK. If buying a lot the PnP from the US becomes less of a significant cost, or have the facilities to reflow them onto boards yourself, but if, like myself, you only want a few and cannot reflow, the XT-Es are much easier to get.
I am not trying to dispute your figures as 1. I don't want to and 2. I don't have sufficient info. But my point still stands that, according to the manufacturers own data sheets, the XT-E does manage to achieve (as a maximum) the same 53% that the luxeon T quotes as typical.
Tim
[Edit] Of course, if what you mean is, "of the bins actually available and on sale" the best of the XT-Es is worse than the worst of the Luxeon Ts, then I can happily believe that. I have just bought some vero 10s as it was so much easier to get decent CRI bins than the alternatives I was looking at...
Last edited:
ReefLedLab
In Memoriam
It is my personal opinion, based on posts of many reefkeepers on many Internet resourses, no more.Not quite sure what you mean by my "opinion is widespread"
I have necessary data. If you would like to know, I can indicate exact data from manufacturer's datasheets. If you don't need this, I would like to say only one thing. Cree - is first of all advertizing company. In day of introducing of MK-R Cree said "200LPW is commercially available now". Since then, year and a half has passed. Until now, is impossible to buy this LED. But many people believe about Cree has a champion efficacy. It is not true.
OK
By the way, you don't need to buy high-CRI LED for your full-spectrum LED fixture. Because you will tune different part of spectrum with different colouring LEDs, isn't it? In this way you can achieve virtually any CRI for any CCT. Unfortunately, high-CRI LED has a bad efficacy.
mussel and hate
New member
I'm not a fan of high density cluster designs such as the lumia or radion. I prefer packing as many 20mm stars as I can into the HS area for optimal blending. I run mine at low drive current to achieve maximum efficiency. The efficiency gain from low current/low temperature operation trivializes the differences in claimed efficiency between manufacturers. It is expensive to double or quadruple diode count though.
Regarding Cree vs Phillips vs Bridglelux vs Semi etc. Once again I'm bucking the trends. I use emitters from each company. There are subtle color differences that I can detect with nothing but my eye between the light from various diodes rated for the same color. This exists even between different batches of the same(supposed) bin from the same company. Diversity within the desired range is a good thing IMHO, at least if properly distributed. It isn't difficult to connect a couple batteries to each diode for inspection and grouping purposes.
Regarding Cree vs Phillips vs Bridglelux vs Semi etc. Once again I'm bucking the trends. I use emitters from each company. There are subtle color differences that I can detect with nothing but my eye between the light from various diodes rated for the same color. This exists even between different batches of the same(supposed) bin from the same company. Diversity within the desired range is a good thing IMHO, at least if properly distributed. It isn't difficult to connect a couple batteries to each diode for inspection and grouping purposes.
reefgeezer
Active member
I already have the 3w Cree's, heat sinks, fans, pots, power supplies, and etc. from my original DIY. The fixture uses 4 dimmable Meanwell ELN-60-48D Drivers. The fixture has ran great ever since I built it. The LEDs aren't old enough to justify scrapping them so I'll use them for the upgrade. I do want to reorganize them as much as possible into "clusters" during the upgrade.
I have read some threads on RC and googled "Audrino", but I'm feeling pretty lost. I'm didn't think I was a complete dunce, until I started researching this project!
I need a boat load of help. I know this is a broad question, but I don't know enough yet to ask better ones... Can someone maybe provide a list of hardware needed to build a system to run somewhere between 48 to 60 3W LEDs that controls time on, ramp up, dwell, ramp down, & off on three or four channels. Even links to good threads would be helpful.
Thanks in advance for the help.
I have read some threads on RC and googled "Audrino", but I'm feeling pretty lost. I'm didn't think I was a complete dunce, until I started researching this project!
I need a boat load of help. I know this is a broad question, but I don't know enough yet to ask better ones... Can someone maybe provide a list of hardware needed to build a system to run somewhere between 48 to 60 3W LEDs that controls time on, ramp up, dwell, ramp down, & off on three or four channels. Even links to good threads would be helpful.
Thanks in advance for the help.
Similar threads
