Best Tool for the Job

This LED forum has been accused of dogmatic unthinking parroting. I see it as a collection of very intelligent and helpful people sharing ideas and information freely, advancing the tech as best we can. But we are human, so put us together and naturally there is going to be friction. Lets use that energy to continue improving lamp design and Please TRY to avoid personal attacks, for the sake of team Grow

There are many different styles of DIY lamp here in RIU. I cannot think of two that are identical. No one is suggesting there is only one way to go. The beauty of DIY is that you can customize it for your specific goals. There are many interesting design goals intensity, efficiency, spread, up front cost, simplicity, aesthetics etc. That said, it is possible to overlook the best available tools to achieve your goals because this tech is changing so fast. That is where we can team up to help each other out with collective knowledge gathering and peer review.

On of the most common questions that comes up, and is a darn good question to be asking, why should we pay more for higher efficiency? A valid idea is reproducible and able to stand up to scrutiny/testing so it is good to question that practice. The law of diminishing returns does kick in at some point and depends on the application, so cost analysis is not always straight forward.

Some basic options
1000W DE HPS = $450->320W PAR W exiting the reflector. $1.40/PAR W all in cost for 40% efficiency
(1) Vero29 @ 2.45A = $63->34 PAR W = $1.85/PAR W all in cost for 36% efficiency
(3) Vero29 @ 1.4A = $188->62.7 PAR W = $2.99/PAR W all in for 41% efficiency
(5) CXA3070 3KAB @ 1.05A = $325->88.7 PAR W = $3.66/PAR W all in 46.7% efficiency
(8 ) CXA3070 3KAB @ .7A = $481->98.8 PAR W = $4.87/PAR all in 50.4% efficiency
(11) CXB2530 3KU2 @ .5A =$380->94 PAR W = $4.04/PAR W all in 49.1%


Why I pay more up front for efficiency

Not long ago I paid $17.80/PAR W all in for early DIY LED at 26% efficiency. Due to the high cost, when it cam time to scale up I went with HPS. But now, paying $4.87/PAR W for high efficiency is fine by me, especially if I am using it to push the limits of how much I can yield in a given space.

Here are some of the potential benefits of increasing efficiency that might interest a light designer:

Reduce system heat, reducing ventilation power, noise, improving stealth and working conditions
Reduce or avoid air conditioning, especially in warm climates/seasons
Increase intensity and yield in a given space
Less building materials required to do the same job (drivers, heatsinks, wiring)
More affordable to use passive cooling
Increased carbon contact time, less carbon needed and/or carbon lasts longer
Grow off the grid
And the most obvious, save on the cost of electricity which varies a lot depending on your location.

I do care about pollution so I try to keep that in mind and be efficient wherever possible.

Very nice write up!!

I just invested up front for geothermal and solar for the same reasons

I tend to disagree with an aggressive pursuit of efficiency today. Now, if you are fighting heat issues, none of this applies as removing heat can be the single most important item to consider. But, the rapidly advancing state of the art in solid state lighting makes being on the bleeding edge terribly expensive. Don't get me wrong, I love to see what is possible and the builds here are awesome. I just can't justify the higher costs of the ultra efficient Crees.

When Newark had their 20% off sitewide sale last month, I stocked up on Vero 18s for $10/ea. The drivers and heatsinks that go with them, will be usable for many generations of new chips and upgrades are just a quick swap away. I would not skimp on the supporting hardware as there will likely not be advances that makes any of that obsolete soon. But the chips are different, while I likely won't replace them for the V2.0 veros, next years version, or possibly the year after that, will be sufficiently advanced, that upgrades will make too much sense to avoid. So the longevity of the chips being run a bit harder is not likely to become an issue. So then the question becomes, how much savings in electrical use will be had over the 3-4 year useful life of the CXB chips, versus the 2-3 years I expect to use the Vero 18s.

Unless you have very high electric costs, exceptional heating issues, or are fairly certain that you will not look to replace the chips with newer models as they arrive each year, I think buying the bleeding edge unnecessary.

I don't buy a new computer every year, and similarly not planning on upgrading my LEDs. No point.

Otherwise I'm looking for long term use and efficiency. A return on my investment in electrical savings. I did my whole house in them

Supra, we think alike. I'm all about big yields in small spaces and with LED it's become a reality for me. I too was an early adopter with led spending way too many dollars per par watt early on. It's been a good learning experience. The way we use LED is as diverse as the people here and I'm no exception. I like to totally surround the plants with light with sort of a light chamber while others like scrog or sog. I really don't think there's a wrong way to do it unless the plants die.

I do have to agree with getsoutalive that riding the leading edge can be a money pit. Years ago I got into building my own computers. Over a period of 5 years I spent nearly $30,000. I had some real cutting edge stuff for a month or two at a time. The lesson for me was that riding the leading edge can cost 4x as much as buying last year's best electronics.

@suprasl you might add that your comparison is always 1 driver with it's ideal number of cobs, otherwise it looks a bit confusing...

Rrog said:

I did my whole house in them

Click to expand...

Me too!! Woop! Props on geothermal though. Hell yea. I don't plan on upgrading my cxa ABs for some time, till there are a little more drastic of differences. But to build a new panel for another area- I want the top of the line (of today) so I won't feel the need to upgrade tomorrow.

I think efficiency goes with the ppl that aren't looking at upgrading the minute their outdone, run smaller spaces/heat issues, etc etc yada yada..

way I see it, Get a dimming driver that has the max current yoy want and/or Pot. So the options their just in case.

I was actually planning on making my third light out of vero 18 again, but cxb had just come out, and only added 100 dollars to my total design to upgrade from 40% to 49.1%.

Vero 18 1.2 was not coming down in price, so I jumped on the cxb2530. Since then, there's been a shitstorm.... so i just assume people do not like that I paid 500 something for my lamp..

I've been acused by just about everyone for not calculating amortization costs, but I was actually doing the amortization vs vero 18, not hps! lol In other words, I was doing a rough amortization cost on how long it would take to pay back the 100 dollar difference in upgrading from vero 18 to cxb2530 u2. The time was about 1 season's worth of growing under the lamp. After that, my vero 18 to cxb2530 upgrade paid for itself.

--- separate train of thought ---

What doer doesn't realize is that even real engineers avoid extremely difficult math problems, and use shortcuts whenever possible. Engineers usually do not spend time solving differential equations. Instead, they think back to the one time they solved the diff eq in college, and then look for the algebraic solution so they can begin designing around that instead.

We've been accused of not showing math in the context of heat sink analysis. The EEVBlog guy said it himself (i tried to say it a number of times here, but people just think I'm being patronizing and condescending when I do). Heat sink analysis is so hard, it's not even worth trying to do more than a rough estimate..

I'm patiently waiting for the ver 2.0 vero 29's to come out. Driving them at ~1amp yields 46% efficiency, ~$4.34/parw on final assembly for my designs. I have people begging to beta test

churchhaze said:

I was actually planning on making my third light out of vero 18 again, but cxb had just come out, and only added 100 dollars to my total design to upgrade from 40% to 49.1%.

Vero 18 1.2 was not coming down in price, so I jumped on the cxb2530. Since then, there's been a shitstorm.... so i just assume people do not like that I paid 500 something for my lamp..

Click to expand...

I think your latest light is great! I appreciate seeing what people build, especially the real good ones. Efficiency is no small matter. It's one of the main reasons why I don't use epiwhatever based lights anymore. I just got an email from my power company. My consumption is 9% greater than the average for my neighborhood. As far as I know, I'm the only one with an indoor garden. I could get it a little lower by spending a lot more at this point. In another year or so I'm sure there will be more efficient emitters available.

Mechmike said:

I think your latest light is great! I appreciate seeing what people build, especially the real good ones. Efficiency is no small matter. It's one of the main reasons why I don't use epiwhatever based lights anymore. I just got an email from my power company. My consumption is 9% greater than the average for my neighborhood. As far as I know, I'm the only one with an indoor garden. I could get it a little lower by spending a lot more at this point. In another year or so I'm sure there will be more efficient emitters available.

Click to expand...

At this point, I'd make a more efficient version just to piss off the jealous trolls.... Some people just hate it when you have "better" stuff. If someone comes to me with an even more efficient lamp than mine, I will know the future is even brighter than I had originally anticipated!! One day, I will also have that efficiency!

I can't wait to get sidelighting up in my tents to boost the 12 PARW/sqft to 15PARW/sqft. 12 is perfect for the top imo, but because I watch what you guys do also, I'm starting to think I could pack in more total light.

I honestly can't believe there is a group that doesn't like what we're doing... Do you have any idea how useful it is to be able to say "look at @Mechmike 's design to give you an idea of how one person does it?" Where do they think ideas come from? Your design alone has given me so many ideas, but I want to change it up a bit.. I'm thinking of using the 12" profile at short lengths to make the side lighting passive with vertically oriented fins.

This RIU sub-thread on LED is far and away the most definitive, most active, and on the cutting edge thread of it's kind that I've found anywhere. I post that all over the place.

Sometimes the required passion to propel such a forum is viewed as arrogant or myopic to the uninitiated.

Rrog said:

This RIU sub-thread on LED is far and away the most definitive, most active, and on the cutting edge thread of it's kind that I've found anywhere. I post that all over the place.

Sometimes the required passion to propel such a forum is viewed as arrogant or myopic to the uninitiated.

Click to expand...

I couldn't agree more.........and supra is a big part of that, no doubt.

Rrog said:

This RIU sub-thread on LED is far and away the most definitive, most active, and on the cutting edge thread of it's kind that I've found anywhere. I post that all over the place.

Sometimes the required passion to propel such a forum is viewed as arrogant or myopic to the uninitiated.

Click to expand...

Perfectly said!