Brotherjericho's stupid questions about DIY and such

personally i see no reason to not include deep red. Its another wavelength which is photosynthetically active and zero risk of oversaturation/bleaching.

edit: misread, oops, personally i tihnk you need to add more deep red, but certainly not "all" deep red. The goal here is to spread your spectra, IMHO. i would want similar amounts of red/deep red.

Also to clarify its my deepreds that i run at 1A. I dont even use monochromatic reds- i use warm whites for that area.


edit2: 6 red/4 deepred would be my suggestion personally, since your WWs will be contributing to the "red" area too

SnotBoogie said:

i run my LZ-1 reds at 1000ma. :3

With adequate cooling its fine to drive at max current. Yes im sure it will shorten their lifespan a little, but i dont think its significant. Cooling is the main thing.

Click to expand...

Yeah its hard to find drivers in 500-550ma range anyway unless I go dimmable. There is 630ma (which I am using as a reference in my graph) but how much is the trade off?

SnotBoogie said:

personally i see no reason to not include deep red. Its another wavelength which is photosynthetically active and zero risk of oversaturation/bleaching.

Click to expand...

I do include it, but to the exclusion of red altogether? Some are suggesting it, but how many people are doing it with good results on several strains? It seems Apache, Astir and Area51 all leave it out as a targeted spectrum.

Playing around with the ratios, focusing on the reds....

18 WW: 2 CW: 6 R: 4 DR



18 WW: 2 CW: 4 R: 6 DR



More deep red seems to allow me to drop a couple of CWs for WW.

Hey Brother J you shouldn't worry about driving CREEs hard if your cooling is there. Driving Chinese mystery diodes at 700mA is not too ....cool. Sorry.

FranJan said:

Hey Brother J you shouldn't worry about driving CREEs hard if your cooling is there. Driving Chinese mystery diodes at 700mA is not too ....cool. Sorry.

Click to expand...

What about the XT-E whites? I can get some extra oomf if I drive those at 1.05A, they are max rated at 1.5A.

turnip brain said:

XTE look like really good whites, but probably not much extra bang for your buck running high. Efficiency drops as amperage rises. supraSPL wrote pretty extensively about the efficiencies of the XT-E: https://www.rollitup.org/led-other-lighting/590648-diy-led-220w-cree-xte.html

Click to expand...

Yeah, I did see that but I'm not certain what would be the trade-off. I know there would be more heat to manage, But I'd also gain almost 1100 lumens at 1.05A over .7A on the whites for a 30 piece cluster, with an accompanying increase of several intensity percentage points in all regions.

brotherjericho said:

I like that Makersled, saw it first a day or so ago. Expensive though, and looks to be only 4" wide. So I'm probably going with 18" of this one: http://www.heatsinkusa.com/8-460-wide-extruded-aluminum-heatsink-with-t-slots/ . Don't get all of the extras but its twice as wide and ~$57 for the size I want.

Click to expand...

I bought 2 of these heatsinks with t slots from HSUSA.

Well made..they forgot to send the extra screws so I emailed them and they answered right away and overnighted
them to me at no cost.

Great service !

PS: I had to snip off a tiny bit of screw because they were a tad too long for rapid solderless leds and I even used the thermal paste...I don't know if they will work with ledgroupbuy because I don't have lEDGB leds. If anything order screws from led group buy also (if you buy their leds) you might save time having to snip usaheatsinks screws if they work out.

Well if I was building a panel for vegging and flower, I might go all CW, and adjust red accordingly. Or go NW and reds. But this is going to be flower only so the focus is to pump that big red curve into the distribution. It's not like plants only 630nm and 660nm, they can use alot more so the warm whites give you that. I need only enough blue to control stretch (hopefully) help with sativas.

turnip brain said:

If I am reading you right, it is an increase of 30% in electrical consumption. Your fixture with 30 LEDS goes from approx 65 to 100 watts output. On the cree flux/amperage graph looks like close to 30% increase in flux, so seems like pretty equal bang for the buck, but efficiency is reduced with higher jct temps, so probably do have to de-rate that flux increase a bit.

Click to expand...

Actually, the figure for 60 pcs (the whole thing) is going to be 114w with whites at .7 and 156w with them at 1.05. The lumens go from 7976 to 10308, so I would lose some efficiency (5.6%). Plus I'd have more heat and one of the purposes of this build is to reduce my heat. I'm betting that at 114w my panel will blow the socks off my PLLs. It better.

Instead of driving myself crazy I went with 2 series..one for the cooler leds and one for the warmer leds.

I have both series on their own dimmer and adjust dimmer accordingly manually and can get any combo of spectrum I want. I'm buying the controller for ledgroupbuy which is made for my driver /dimmer so I will be able to time it thru out the day and not have to worry about doing it manually anymore.

Saves time worrying about if you have too much or too little blues or reds when vegging or flowering..just adjust dimmers on each series to dial in whatever you want or get a controller with timer to mimic what you want.

Heat isn't an issue if you have a big enough heatsink and a fan.

turnip brain said:

So at 5.6% loss of efficiency, is there a long term cost benefit of running at higher amperage vs adding more LEDS?

Click to expand...

That I don't know, but I'd rather just keep things at .7A and see how it goes. All I want is better heat management than I have now, but at least a little better performance. I'm waging that 114w of Crees and Rebels on a cooled heat sink will be much better than 156w of PLLs. I would like to get my temps down to the high 70s, low 80s. Right now it hovers around 85F and I want to drop that down.

Next dumb question: how do I use an "Effective Spectral Distribution" chart beyond just seeing that I have what I want, visually? For instance, here are two recent charts that were developed using http://ledspecs.com/



The first is a mix of 16WW:4CW:8-630nm:2-660nm. The second is the Hans Panel. Visually, they both have similar curves and intensities in the blue and red region. The custom panel has 4624 lumens, and 17333mW total watts. The Hans has 3271 lumens and 22270mW. So what can I glean from this infor?

That your panel, before overhead, is better than Han's in fantasy land .

BrotherJ do yourself a favor and hit up BitTorrent and look for this book. (Click on the pix for the Amazon link if your conscious gets the best of ya )

This book really opened my eyes to the problems of making lights and how decisions regarding LED design are made. I like this book especially for the more real world design they talk about. You seem like someone who would read this and gain some things. There's also a few other solid textbooks on LEDs available on BT including Schubert's, which most people should try to read or have for reference.

FranJan said:

That your panel, before overhead, is better than Han's in fantasy land .

Click to expand...

It is pure fantasy! But what I'm looking for is a way to figure out percentage of blue based on that info. I know the intensity but that is not the same thing. I want to make sure I get roughly 15% blue in my overall mix, and do it without actual blue chips.

BrotherJ do yourself a favor and hit up BitTorrent and look for this book. (Click on the pix for the Amazon link if your conscious gets the best of ya )

This book really opened my eyes to the problems of making lights and how decisions regarding LED design are made. I like this book especially for the more real world design they talk about. You seem like someone who would read this and gain some things. There's also a few other solid textbooks on LEDs available on BT including Schubert's, which most people should try to read or have for reference.

Click to expand...

I need the other book "LED Design for Dummies".

Thanks!

Still debating the white issue in my DIY project...but now a different debate. I'm looking at two different choices on WW: Philips Rebel ES or Cree XT-E. For efficiency, it would be the XT-E all the way. But (there's always a but), look at the two spec curves.

Rebel ES:



XT-E:



I really don't know how to calculate efficiency on LEDs yet, but I do know the Rebel ES has a typical luminous flux of 135lm @ 700ma while the XT-E has quite a bit more at 167lm+ (based on bin). However, its 135lm of a better grow distribution vs 167lm of a decent one. Suggestions?

I don't know that you can lose choosing either Cree or Philips generally but between the two, I would opt for the Cree for white...but I keep choosing the Luxeon ES for deep red.

When I became interested in LEDs a couple of years ago, I took a short course from Cree on LED lights where we actually built a fixture. Their design choices but each choice was figured and explained along the way. I enjoyed it, learned a lot and have been dabbling since. I think that I'll check out the book that FJ is recommending. Who knows what other jewels of wisdom, I might glean.

Yeah, I'm at a loss of choosing between efficiency and SPDs. If I decided to ever try a grow with only one color white, I'd use the Rebel ES 3500k.

Well, I think after much playing around and taking into consideration my set-up for veg and such, I have decided to go with a combination of whites and reds after all. I'm not concerned about making the g/w list anywhere, just managing my heat in my flower tent and growing successfully with as few hiccups as possible (from my lighting). That being said, I am going to build a 60 piece panel, 30 per targeted plant. I will run this with 4 drivers, 2 for each thirty piece. That way, if I'm only growing one plant, I can switch off one section altogether. This is my ratio for each section, all at 700ma max:

8 - Cree XTE WW
12 - Cree XTE NW
4 - Cree XPE Red
6 - Philips Rebel ES Deep Red

My plants are all vegged under 144w total of 6500k PLL lamps, to keep them short and bushy. Once I put a plant in the flower tent, I will put it under a section with only the whites running. This should reduce the shock by reducing the blue moderately while not adding much red, only that which comes from the NW and WW. This should also minimize stretch. Once I see stretch is over, I will ramp up the reds, which will be on a dimmable driver. Some time during flower, I will turn the reds all the way up, not sure when yet but I have plenty of time to play around.

That's what is cool about DIY, I can tailor the light to my needs. After much research, I hypothesize that this is the best set-up presently.

Thoughts?

BroJ- have you looked over inda-grows red pantoons. If you look on the product page there is a link for more info. I will shorten for here. By leaving only 730 reds on for 5 mins after lights out, mimics sunset and speeds up the change flowering response. On my panel I am going to add a plug for my reds only. That way I can turn them on 10 mins before lights on and 10 mins after lights off, to replicate dawn and dusk. This is what indagrow said, it will change flowering response from 2 hr. to 5 mins. Of coarse this info is coming from a company that is trying to sell a 40 watt $750 supplimental light

^^ Thanks, but not going to bother with 730nm stuff right now. I know the concept, just not ready to move ahead to something beyond focusing on this DIY panel for the moment.