Why are COBs better, and whats so great about the 3590?

[optzulu]

So would you guys say efficiency is more important than spectrum?

led got picked up over HPS when the efficiency start to shine. Not the spectrum quality.
I am going to weight my dry buds today I had 600gram last run from 400W of cxb3590@1400. But I can tell ya Going to hit the 650g for shure. Quality nugs only becouse we sell this stuff. Sorry sometimes a forget a digit so it could also be 6500g haha

 

[matriculated bean]

So would you guys say efficiency is more important than spectrum?

I'd say it depends on things like how much space you are lighting and what you pay for power and how much having less heat factors into your situation and to a lesser extent bulb replacement. If you are replacing a lot of bulbs though it's going to cost you that much more to switch to LED. I think LED spectrum and efficiency will continue to improve faster than HID though. .02

 

[welight]

I've read that because the lumen output doesn't follow an expected curve and deviates over time that the longer they run tests the higher the LM end up being, with some samples showing LM70 at over 200,000 hours.

View attachment 3778702

So much of the data produced is probability(expected) data, I believe as the real performance data rolls along it keeps shifting the probability to better numbers. I recall when XPG released some years ago the early long term summaries were listed at approx 80-90K hours, the last time I looked they were at like 496,000 hours, based on around 30000 hours of actual running data
Cheers
Mark

 

[CobKits]

the last time I looked they were at like 496,000 hours

thats acceptable...... i guess

 

[rob333]

Wrong his name is Johnathan goldsmith he is 77 years old and alive and well..

Feel free to quote yourself spouting drivel.

well the meme been posted that much he is dead to me lol

 

[TacoMac]

So much of the data produced is probability(expected) data, I believe as the real performance data rolls along it keeps shifting the probability to better numbers. I recall when XPG released some years ago the early long term summaries were listed at approx 80-90K hours, the last time I looked they were at like 496,000 hours, based on around 30000 hours of actual running data
Cheers
Mark

The key word in this is: projection.

It's just like stock companies do. They take a performance level they HOPE to achieve and inflate it up into the future to build projections, which make themselves look better, which makes their stock price increase.

60% of the time it never lasts because inevitably they can't meet the inflated projections.

Basing the performance projections of a light 56.5 years into the future based on 3.5 years of running is, at best, a ridiculous standard to even offer up, let alone publish.

 

[Airwalker16]

The key word in this is: projection.

It's just like stock companies do. They take a performance level they HOPE to achieve and inflate it up into the future to build projections, which make themselves look better, which makes their stock price increase.

60% of the time it never lasts because inevitably they can't meet the inflated projections.

Basing the performance projections of a light 56.5 years into the future based on 3.5 years of running is, at best, a ridiculous standard to even offer up, let alone publish.

COBs are INVINCIBLE, BRAH!!

 

[TacoMac]

Of course they are. Somebody spent money on them, so they must be.

 

[CobKits]

The key word in this is: projection.

It's just like stock companies do. They take a performance level they HOPE to achieve and inflate it up into the future to build projections, which make themselves look better, which makes their stock price increase.

60% of the time it never lasts because inevitably they can't meet the inflated projections.

Basing the performance projections of a light 56.5 years into the future based on 3.5 years of running is, at best, a ridiculous standard to even offer up, let alone publish.

yet there are industry standards for predictive failure in every single segment of the electronics industry

 

[TacoMac]

yet there are industry standards for predictive failure in every single segment of the electronics industry

I have never in my life seen one that predicts 56 years into the future. Typically, by the time 10 years goes by, the technology you're using is redundant.

It's all meaningless puffery and nothing more.

 

[PhotonFUD]

they kind of are. for a given chip model, efficiency measured in lumens per watt will always be inversely related to QER

You can still apply both if you don't limit yourself to a single chip. For example, pick your spectrum. It can be comprised of one white or a combination of whites and monos. One chip model or many.

For efficiency that can be addressed with....more chips under driven further!

That is what I meant by mutually exclusive. Both have different approaches that can be applied to achieve similar goals.

 

[CobKits]

I have never in my life seen one that predicts 56 years into the future. Typically, by the time 10 years goes by, the technology you're using is redundant.

It's all meaningless puffery and nothing more.

mmmmm... puffery.

meanwells are rated over 100K hours, btw.

 

[TacoMac]

mmmmm... puffery.

meanwells are rated over 100K hours, btw.

Just goes to show you're once again spewing complete junk.

There's just a little bit of difference rating something for an 11 year life expectancy and performance based on a 3 year trial run, and rating something for OVER HALF A CENTURY based on that same limited data.

You really should stop entering into a battle of wits completely unarmed.

 

[CobKits]

yes sir.

think for a second about a manufacturer with a decades-old understanding of diodes, and after tens of thousands of hours seeing a linear trend of say 98-99% lumen maintenance, that you could predict LM70. its really not that hard of a concept.

 

[PhotonFUD]

This might help:

https://en.wikipedia.org/wiki/Mean_time_between_failures

 

[JorgeGonzales]

So much of the data produced is probability(expected) data, I believe as the real performance data rolls along it keeps shifting the probability to better numbers. I recall when XPG released some years ago the early long term summaries were listed at approx 80-90K hours, the last time I looked they were at like 496,000 hours, based on around 30000 hours of actual running data
Cheers
Mark

Where are you getting these numbers? I don't know what top secret internal files you have access to, but this is what they tell the plebs:

XP-G:

 

[welight]

Where are you getting these numbers? I don't know what top secret internal files you have access to, but this is what they tell the plebs:

XP-G:
View attachment 3780054

yep, would not say they were secret, just not published, Cree has reams of long term summary data, what they choose to reveal in public data sheets is I guess a marketing decision. To be honest I am not sure I would publish that data either, a led that lasts a half million hours, what use is that, who will live long enough to verify and does anyone care. I am interested in the fact they have the technology to do it and that kind of tech feeds into all their products and it tells me what data is published is rock solid
Cheers
Mark

 

[THE KONASSURE]

you mean the phosphor resin coating??

I guess that must be it, looks like a layer of silicone to me but sure, you mean the stuff that`s got a blue tint on cool cobs and a yellow tint on warm cobs and is normally clear on blurples ?

 

[CobKits]

that is the phosphor layer which gives the cob its color. if youre gonna peel it off just save yourself the trouble and buy 6500k cobs.

i cant imagine the spectrum youre left with after you peel off the phosphor is very ideal for growing plants

 

[JorgeGonzales]

i cant imagine the spectrum youre left with after you peel off the phosphor is very ideal for growing plants