Custom LED Canon

Well I plan on running the vero 10 at possibly 700ma which will be 1620lm, and as per Jedi a 1:14 ratio looks good. Being I havent seen one in person I am just relying on his opinion. Running the 4 luxeon m's at 1 amp will give me about 24840mw which is very near a 1:14 ratio...and I have no doubt he will pop up soon enough.

He's pretty much on the target (that basically gives you the old school 1 white to two royal blue ratio), though with the added lime and cyan you may need more red since the vero's already cover the 480 to 530nm range pretty well.
 
My eyes have seen it but I cant share without Bridglux express permission. It takes a bit of persistance to get a hold of the spectral graphs for the 90CRI models. The 97's simply have a larger percentage of Red phosphor in relation to the other spectrum. The 97 CRI data sheets are publicly available.

The 90 CRI 5000k for example is pretty well balanced for greens and red phosphor for reef use as a main white source, but with the addition of cyan and lime you may(possibly) end up wanting more red from the vero to balance out (fortunately they are only a few dollars and will be easy for you to swap out if needed.

Alternatively one might add 3 reds instead of limes and combine limes and cyan with 2 limes and 3 cyan (or visa versa) on that channel to give a little better control of overall color temperature regardless of the Vero chip used. I'ts really just a balancing act and much is up to personal preference in the end.

So the datasheet does exist...lol

This are all very good points, and I believe it will just come down to running the setup and figuring what I want. I think it will take me several MCPCB boards before im happy with the led combos and ratios. So I believe I will produce one before I produce all 5 so I can play with color combos as needed.
 
He's pretty much on the target (that basically gives you the old school 1 white to two royal blue ratio), though with the added lime and cyan you may need more red since the vero's already cover the 480 to 530nm range pretty well.

I thought the vero's were actually pretty defecient in the 480 to 530nm range. Looking at the spectral graph from the 80 cri vero 10, it is under the spectral half width, and a little better on the 97 cri datasheet. Or am I not understanding the data correctly? I am pretty new to understanding all this info.
 
I thought the vero's were actually pretty defecient in the 480 to 530nm range. Looking at the spectral graph from the 80 cri vero 10, it is under the spectral half width, and a little better on the 97 cri datasheet. Or am I not understanding the data correctly? I am pretty new to understanding all this info.

Your reading it correctly. on the 90CRI models (not published publicly) the low point of the graph at around 480nm is a bit higher relative to the other wavelengths. just under 30% relative to the other wavelenths in the chip. It is still low but much better covered than most LED chip data sheets I've seen.

Adding blue is going to be good but adding too much cyan or lime will over power the red in the chip.

there is a lot of balancing that has to take place you can either start cool and add red to warm up or start warm and add greens to cool the light. both methods achieve the same effect overall.

I was merely suggesting that the vero already covers the green/cyan/lime range very well and adding more may over power the red present and make things look weird.

Additional 480nm is still a good supplement choice.

the green/cyan/lime to red ratio is the tricky part.

this is why I suggested the decor versions might work better on your array, but you will have to play with it to find out. fortuately the veros are cheap and you don't need many.

I like the graph on the 5000k 90 CRI because it looks much like that of natural sunlight a few meters below the water, aside from the large royal blue spike and lack of 480.

you just need to fire it up and play around a bit. tinkering is half the fun anyway. right :)

I keep several montipora that need that red spike to look a nice bright pink/red color otherwise they tend to look more blue purple, so for me I favor some additional red in the spectrum for that purpose.
 
Last edited:
Your reading it correctly. on the 90CRI models (not published publicly) the low point of the graph at around 480nm is a bit higher relative to the other wavelengths. just under 30% relative to the other wavelenths in the chip. It is still low but much better covered than most LED chip data sheets I've seen.

Adding blue is going to be good but adding too much cyan or lime will over power the red in the chip.

there is a lot of balancing that has to take place you can either start cool and add red to warm up or start warm and add greens to cool the light. both methods achieve the same effect overall.

I was merely suggesting that the vero already covers the green/cyan/lime range very well and adding more may over power the red present and make things look weird.

Additional 480nm is still a good supplement choice.

the green/cyan/lime to red ratio is the tricky part.

this is why I suggested the decor versions might work better on your array, but you will have to play with it to find out. fortuately the veros are cheap and you don't need many.

Got it...thought that i may have been reading it incorrectly. But yes I have been told more is less with the blue and cyan. Part of the reason I tried to seperate the colors the best I could to tune accordingly. But I am thinking of ordering the 97cri also and running one each of the 90 and 97 side by side to see the difference.
 
Got it...thought that i may have been reading it incorrectly. But yes I have been told more is less with the blue and cyan. Part of the reason I tried to seperate the colors the best I could to tune accordingly. But I am thinking of ordering the 97cri also and running one each of the 90 and 97 side by side to see the difference.

might as well the vero's are cheap and swamping them should be very easy. Try out a few different ones.

There will come a time soon when you can just use color LEDs to replicate a traditional MH or fluorescent spectrum and only use a little bit of phosphor converted white to fill in the rest.

the reds are nice and narrow on thier spectrum graphs but so far the greens and lime are pretty broad which is not ideal.
 
Well after several months with no updates I finally received the finished boards in the mail. Here are the pics below....now for the next step since I have the boards in hand. Time to design and manufacture a pendant style heatsink for this beast.

photo-2_zpsd12d39be.jpg


photo-1_zpsc0aa77d7.jpg
 
Last edited:
Where do the Vero leads solder onto the board? and more importantly is Steve's going to be carrying these for you so the rest of us can get our hands on them :)
 
Right behind the molex header are two solder pads which is where the pico ezmate harness will be soldered. As far as Steve's stocking and selling no they won't...as I paid to keep these under my control. I would have no problem running a group sale for these if there is enough interest, but the cost of the luxeon uv LEDs used will probably scare most away. Just in UV LEDs alone cost is pushing $125 if I remember correctly.
 
Can't wait to see them up and running! How are you planing to route/secure the ezmate leads so they don't interfere with the Luxeon Z? MkII might want to put the solder points on the inside of the cluster of luxeons closer to the Vero chip? Just a thought. Awsome project!

If you don't mind the ask what is your total cost per board all said and done?
 
Can't wait to see them up and running! How are you planing to route/secure the ezmate leads so they don't interfere with the Luxeon Z? MkII might want to put the solder points on the inside of the cluster of luxeons closer to the Vero chip? Just a thought. Awsome project!

If you don't mind the ask what is your total cost per board all said and done?

The overall cost per board...I would have to sit down and figure that out, but I do know that these can be made cheaper if made in a larger run. I also ended up with a lot of shipping charges buying all the pieces from different places. I'll let you know when I figure the cost out.

As far as the wires being routed I'm just gonna use a dap of hi temp silicone to keep them secure and out of the way. I had originally wanted to do what you said but the pads had to be moved back to the current location to allow room for the led traces and keeping the overall footprint right enough to use the optic I designed this around. But hopefully I can come up with something better if I do a revision.
 
So for those curious minds the cost per board not including heatsink, shipping or power supply is a little short of $300.
 
Last edited:
The optic is a collimator lens purchased off of a popular auction site...don't really know what the rules are about posting up sites in the forums.
 
Back
Top