T5s to QBs..

[Randomblame]

Loving the discussion so far, we got ppl for led, strip led and t5. Exactly what I wanted, love hearing everybody's honest opinion and experiences! Thank u fellas

With 4ft. F-strips you could even use the old T5 case to mount the strips. Remove bulbs and sockets and glue the strips to the refelctor using double sided thermal tape. It even offers space for the driver.
Would save the cost for frame and cooling profiles!

 

[eLite_gRowr]

How about let's see some pics of plants vegging under the various light sources??

Great point!! Lets see some pics of vegging plants under strips and boards. @Roger A. Shrubber has t5 taken care of

 

[wietefras]

I was wasn't aware he was asking about uniformity or the cheapest option I thought he was asking what's better performance.

Well that's part of the actual performance. Why would you only look at marketing figures? Look up how much light gets wasted when a board is hung in an actual room at the correct height. After you lose 30% or 40% on the walls for a board, do you still think getting a few percent more light from a more expensive smd is worth it?

"Performance" is not a fixed number. If you have 168lm/W strips for $25 a piece and a 171lm/W board for $50, which one offers the better performance? You would say the boards, because that's what the marketing says. What if you buy say twice as many of the strips (which improves the performance by 10%)?

Like I already explained, even on datasheet figures, strips will beat a QB. Cheaper strips allow you to buy more, run them softer and get better performance. That's why "old" F series strips still equal or beat the Q series with those "more efficient" LM301B smds.

That's even before you waste 10% to 20% more of the light on the walls with boards versus strips. Boards would actually need to be about half the price of strips for them to give better "performance".

 

[Ryante55]

Well that's part of the actual performance. Why would you only look at marketing figures? Look up how much light gets wasted when a board is hung in an actual room at the correct height. After you lose 30% or 40% on the walls for a board, do you still think getting a few percent more light from a more expensive smd is worth it?

"Performance" is not a fixed number. If you have 168lm/W strips for $25 a piece and a 171lm/W board for $50, which one offers the better performance? You would say the boards, because that's what the marketing says. What if you buy say twice as many of the strips (which improves the performance by 10%)?

Like I already explained, even on datasheet figures, strips will beat a QB. Cheaper strips allow you to buy more, run them softer and get better performance. That's why "old" F series strips still equal or beat the Q series with those "more efficient" LM301B smds.

That's even before you waste 10% to 20% more of the light on the walls with boards versus strips. Boards would actually need to be about half the price of strips for them to give better "performance".

Your failing to factor in labor if you want to calculate it like that a whole day of building is worth $300

 

[wietefras]

Your failing to factor in labor if you want to calculate it like that a whole day of building is worth $300

Pffffft. Come one man, now you are really grasping at straws.

Is it really that much more work to put together a COB kit vs a board kit?

Besides, just look up the Ledgardner PPFD matrices. Those show again and again, that over 40% of the QB's light is lost on the walls. With strips or COBs you can half the distance to the canopy and thus cut the wall losses in half. That's a lot of extra light on the plants and indeed well worth a bit of time/elbow grease.

 

[Ryante55]

Pffffft. Come one man, now you are really grasping at straws.

Is it really that much more work to put together a COB kit vs a board kit?

Besides, just look up the Ledgardner PPFD matrices. Those show again and again, that over 40% of the QB's light is lost on the walls. With strips or COBs you can half the distance to the canopy and thus cut the wall losses in half. That's a lot of extra light on the plants and indeed well worth a bit of time/elbow grease.

What? You were talking about building a light with a shitton of strips. Also I have a bunch of boards that I have ran hard an soft there is a point of diminishing return when you run too soft you lose alot of penetration.

 

[Greensea]

A HLG 100 will replace 200w of T5.

 

[Randomblame]

If you already have all the parts, putting them together is done quickly. Drill a few holes for the frame, use thermal tape to attach the strips, wire them in parallel to the driver, hang it to the ceiling and that's it. With two 2ft. double row strips( eqally to a quantum board28 it should be done in 1-2 hours. With only two DR strips you get better distribution, with 4 single rows it would be even better. Same brightness across a 2x 2' with only one powerful board in the center at the same height is impossible. With 4 boards above a 4' x 4' you'll need at least 16", better 20 or 24". With 8 4ft. SR strips the light could hung 6-8" above the plants to get good uniformity. Average PAR readings would be much higher and you could dimm it by a certain marge and still get the same avarage PAR readings like a board.
The efficiency of a lighting depends not only on the electrical efficiency of the light source. Distribution, heights and uniformity are also crucial factors. A 150w COB over a 2x2 'will not do you much good if you only have 12" of height left, no matter how efficient it is. Even the best chip would cause a hotspot in the center.
Better uniformity is one of the main reasons why strips work so good. The darkest area in the corner of my 2x 4' tent gets still 80% of the center measurings. When I set the intensity in the center area to let's say 50klx I would get more than 40klx readings in the corners and +45klx along the sides.(2' away)
With my 8 CXB3590's+ Osram 660's I needed 380w and a distance of 24" to get the same uniformity with no hotspots. Now I need for the same values not even 280w at only 12" distance. And i've yielded the same 50g/sft. as before.
That's not only because the diodes are more efficient, the way how the photones are distributed is much more efficient and much less light is wasted on the walls.

 

[eLite_gRowr]

I do have 4- 4x4 rapid substrates but those are intended for the new flower room(currently running hid) What do u guys think about individual cobs in the veg room just hanging from heat sinks on ratchets?

 

[Randomblame]

Hmm! You can literally get the same light from COB's but the required heatsinks would make the solution a bit more costly.
As I said in my last post, you could glue 4 single row or 2 DR strips directly into the 4ft. T5 housing(100$ on strips) and use an HLG-240H-48A(65$) as driver. No frame, no hangers, no heatsinks, only double sided thermal tape, the strips and a driver is needed and 4 3500 or 4000°k single row strips would replace eight T5's more than adequately. You can even use the single core wire usually used to wire T5's(AWG20) to wire the strips to the driver, lol! Even the wall plug could be reused!
That means the whole conversation for both lights could be done with only 330-350$ for 560w at the wall!!

 

[eLite_gRowr]

INGENUITY!!!!

 

[wietefras]

What? You were talking about building a light with a shitton of strips. Also I have a bunch of boards that I have ran hard an soft there is a point of diminishing return when you run too soft you lose alot of penetration.

No I'm not talking about using a shitton of strips. A 112cm double row strip has the same output as a QB288. So you'd need similar amounts of those led strips as you'd need boards. For a smaller tent you'd need two 2'strips to replace a board. Still hardly a "shitton".

You penetration statement is simply incorrect. You don't get less penetration if the amount of light stays the same. Diminishing returns yes. Indeed there is no real advantage in penetration anymore from using single row led strips over double row ones. However, going from 1 board per 2'x2' to 4 COBs or 2 strips over that same space does improve penetration a lot. First of all from spreading the light better, but also by lowering the optimal fixture height to half the height needed for a QB.

 

[wietefras]

Hmm! You can literally get the same light from COB's but the required heatsinks would make the solution a bit more costly.

I have put COBs on aluminium strips. Essentially you make a "COB strip" that way. Cooling requirements are then no different from a led strip really. I chose to put 2 CPU coolers on each strip, but I also could have used a U channel or length of heat sink.

Cobby put four Citizen 1212's on a cooler intended for a QB. Worked just fine too.

I agree that it's more practical to use led strips though.

On the other hand, I had a lot of old COBs laying around and strips don't fit well in a 100x100cm tent, so I repurposed those old COBs by putting 24 of them on a 3'x3' frame. Wiring was a mess (lots of parallel and serial wiring mixed), but it works great and it's nicely efficient (24 CXB3070AB/AD COBs on in total 367W)

 

[Randomblame]

Good exapmle for a COB-light with equivalent efficiency and uniformity compared to an F-strips build. I bet you need only 12" distance to get even PAR/lx measurements across the canopy...?!
Old COB's can still achieve very good efficiency despite their age, @14,5w / COB you should get ~180lm / w or so, right?
I reuse 6 CXA2540 currently running @175mA to reach the same effiency range, but it's a strip/COB combi for a relative small 3sft area. Although they are already 4 or 5 years old, they still have their uses.

 

[wietefras]

I bet you need only 12" distance to get even PAR/lx measurements across the canopy...?!

For determining correct height I use the rule of thumb "2/3 of the distance between the COB rows". So that's 25cm * 2/3 = 17cm (10" * 2/3 = 7"). With strips I feel you can even go as low as "1/2 the distance between the strips".

Simulation of the light distribution at 17cm (7"):


Only the corners are a bit too dark.

It's always a trade-off between uniformity and losing light on the walls. At 20cm the light is even more uniform, but then 2% more light is lost on the walls. I actually use 15cm as the height. Bit less uniform than at 17cm, but also less wasted on the walls.

Old COB's can still achieve very good efficiency despite their age, @14,5w / COB you should get ~180lm / w or so, right?

It's a mix of AD and AB COBs (with COBs on 11W, 15W and 23W), but according to the datasheet the average should be about 170lm/W or 2.5umol/s/W.

 

[Randomblame]

For determining correct height I use the rule of thumb "2/3 of the distance between the COB rows". So that's 25cm * 2/3 = 17cm (10" * 2/3 = 7"). With strips I feel you can even go as low as "1/2 the distance between the strips".

Simulation of the light distribution at 17cm (7"):
View attachment 4148721

Only the corners are a bit too dark.

It's always a trade-off between uniformity and losing light on the walls. At 20cm the light is even more uniform, but then 2% more light is lost on the walls. I actually use 15cm as the height. Bit less uniform than at 17cm, but also less wasted on the walls.

It's a mix of AD and AB COBs (with COBs on 11W, 15W and 23W), but according to the datasheet the average should be about 170lm/W or 2.5umol/s/W.

.. but really only in the extreme corners, mate, for me that looks nice and even..!
2,5μMol/j is more than acceptable for a 4 years old piece of tech.
F-strips at nom. current(1120mA) are ~2,4μMol/j and ~2,6 @700mA.
IMO, good enough for a few more years!

 

[eLite_gRowr]

Guys thank u for all the kind and supportive information u have provided in the last couple of days, all of it was great! I've learned so much in this post!
I just ordered 2 hlg260 xl qb v2 kits in 4000k and my flower room pucks should be here anyday now
Time to experience led first hand

 

[eLite_gRowr]

Popped some beans to run under the new 260XL... Stand by

 

[Rocket Soul]

You got some very spot on advice so far. One problem with strips is that the sammy strips are sold out by most distributor and not very available al though someone around might correct me.

You may wanna consider bridgelux strips aswell, both EB strips and Vesta strips are economic. Eb Eb stirps have a similar efficiency to Samsung, maybe just alittle less. Also dont need heatsink or anything if you run them at nominal. I like the vestas in particular: they offer something quite unique: dual 90 Cri spectrum 5000k and 2700k, in one strip. You could run a straight 5000k spectrum for vegg or both at the same time, coming to around 3850k which should give you a nice vegg spectrum. Or you can have both powered separately and dial in your spectrum after how plants react.

Cons
- they arent as efficient (as per datasheet) as the samsung strips. The lums per what come out lower, but as per photons they arent that far behind.
- 90 cri seems to stretch a little bit more in vegg than 80 cri. But theres also quite a few who love vegging (and flower) in 3000k 90cri due to the fast growth rates, not just stretch but stalks aswell. At least in what ive read around riu.
- if you find that only 5000k 90 cri works for you then the 2700 diodes wont be used, making them useless.
- no growlogs or testing of these strips available

Pros:
- less than 10$ per 50w strip
- the ability to dial in your spectrum
- 90 cri spectrum; vigorous growth due to added red and far red.
- the ability to dial in your spectrum, controlling stretch/internode distance etc. As stated above: this comes with a cost of efficiency; your total wattage would have to be 10-20% more than with samsung strips for equal amount of photons. But with 90 cri your photons would be "worth" more cause there is more deepred which is more easily used by the plant.


I cant recomend these based on experience as i havent set up my strips yet. But they sure look up to doing the part, especially the "half and half" spectrum. Attached here, in comparison to samsung 3000k, in absolute terms: you see how much more light you get with samsung. But i like the fact that the vesta spectrum is so "flat" that it literally covers every part of the action spectrum, much like a CMH which is said to vegg like a beast.

Ok, long ramble, im sure someone will come up with arguements against them, but is an option

 

[eLite_gRowr]

HLG 260 xl Doing great so far