What's new
TerraForums Venus Flytrap, Nepenthes, Drosera and more talk

Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

DIY COB LED info

  • #81
Happy Thanksgiving!

Another DIY LED update. I made a plastic container based grow space. I used Samsung H series LED linear modules this time. This module was announced in Summer 2016, but it started to ship in October. This Samsung basically beats Philips XF-3535L (I posted about this previously) in all aspects. It has amazing efficacy (187lm/W), even higher than most COB LEDs. I haven't calculated the PAR efficiency, but I'm pretty sure there is no commercially available grow light with this high efficiency. It is a bit more expensive per light output than COB, but it is easier to assemble, and better for a small area. Also, it provides very even light distribution. It costs about $60 to cover 2-3 sq. ft.

I built this for pygmy Drosera, so it is targeting relatively high-light plants. Last year, I was using much weaker light (PPFD of about 70-100 micomol/m^2/s). Surprisingly they did ok. I doubled the amount with this H-series. So we'll see how the plants respond. It takes about 1 hour to assemble.

Here is the link to my Orchid Borealis blog post.

PB130143.jpg


PB120013.jpg


This is supposed to be Drosera dichrosepala ssp. enodes from Scott River. It is with the previous weaker light.
PB130052.jpg
 
Last edited:
  • #82
That's a pretty nice setup! Keep us updated.
 
  • #83
Thanks for publishing this, I've built 2 of them. I'm just getting back into growing again.
f3a303528f70bb4bfce500f704ba8bbd.jpg


I added fans to move the heat away and everything is run by a 10A power supply.
8710ba6cf6aa71f242c44608468e2c75.jpg
 
  • #84
They look very nice, Dave. Are they Samsung H? Did you attach them to aluminum cookie sheet or something as the heatsink?
 
  • #85
They look very nice, Dave. Are they Samsung H? Did you attach them to aluminum cookie sheet or something as the heatsink?

Yes, I used the LT-H282D modules and attached them to aluminum U channel with heat sink compound. I built frames from left over cedar lumber and FRB wall board.

The one on the left has been running about 2 weeks and the VFTs and Nep are really starting to get nice coloration.
 
  • #86
I'm about to take the plunge and buy some of these new LED strips to grow lowland Nepenthes (and other plants later) in one of those mini greenhouses. Interestingly, the shelves are about the same size as the tub in your blog post. Right now I'm trying to decide between 4x1ft (will be mounted slightly lower due to shelf support bars, with more air above) or 2x2ft modules (mounted slightly higher, no air gap.) I'm also unsure about the drive current since I know that lowland Nepenthes don't like a lot of light.

Any suggestions?
 
  • #87
Thanks for the info about the frame, Dave.

Nimbulan, 2ft x 2 is probably cheaper, and less connection means that it is easier to assemble. If you have many shelves, you might be able to go above the quantity required for the volume discount (this is the reason I got 1ft strips). I don't pretend to know well about the light requirement for Nepenthes. I have been growing about 10-20 lowland species, and they are sharing the space with my orchids, they seem to be growing ok (about 100-150micromo/m^2/s, I think). If your shelf has the similar dimension (including hight), putting 4x 1' strips per shelf is probably at the higher end for Nepenthes. So 3x 1' strips might be appropriate (or 2x2', but at a lower current).

Is the distance between the light and top leaves about 1'? How many shelves? Is it ok to drive all LEDs by a single driver? This is more economical, but you have to turn on/off all shelves simultaneously. What's the approximate cost of electricity ($/kWh)? With these info, I might be able to suggest an appropriate combination.
 
  • #88
I would say it's right about 10" between the top of the pots and where the LEDs would be mounted if I use the 2 foot strips, and 9 inches with the 1 foot. Each shelf is approximately 27" x 18". There are 5 shelves and the top one has a domed top above it so I could potentially mount the LEDs higher there. I wasn't planning on setting them all up immediately - I really don't have many plants to put in there at the moment. I would like to set aside one for petiolaris sundews and some lowland-type Utricularia like U. fulva so that shelf will need significantly more light.

I had one concern though - in your original blog post you set up 4 LED strips, wiring pairs in parallel. If one of the strips failed, wouldn't you be risking destroying the other one with a constant current driver or is there some kind of failsafe built into the LED strips?

Edit: I should also add I would like to replace a T5 fixture I have set up on a 36"x13" shelf, again with approximately 10" of vertical space (can be adjusted.) I'm using a 2' fixture right now but it doesn't provide very good light on the ends of the shelf. This shelf will need its own driver.

Edit 2: It looks like Arrow has some 2 foot modules on order and like usual their prices are WAY better than Digikey. They may be reserved, but if they aren't I'll be buying from there in a couple days.

Edit 3: Also do you have any other recommendations for thermal glue? Amazon doesn't seem to sell that product anymore and the shipping time from Deal Extreme would be quite long. I can get the same product from a US seller on eBay but it's $14 though there are several other similar products such as this suspciously cheap one 1pcs Curing Silicone Thermal Conductive Grease Paste Compound Sticky Glue 60g | eBay one that is similarly-sized 50ml - Thermal Conductive Silicone Plaster Glue Adhesive - LED GPU Heatsink | eBay and some smaller ones Large* 10g -Thermal Conductive Silicone Glue Adhesive - LED GPU Heatsink Mosfets | eBay
 
Last edited:
  • #89
With the higher light plants, I think 2x 2' strips is probably a good start, and I think Nepenthes will be ok, too.

You are correct if one of the LED modules completely fails, the other will get 2x current, which is probably close to the maximum current. I used to worry about it. But that kind of failure requires that the wire get disconnected, which is unlikely to happen. Or 3 diodes dies simultaneously, which is unlikely. Within each module, there are several parallel circuits. In case of 1' strip, there are 3 parallel circuits per module. So if 1 LED dies, then 1/3 of the diodes will be dead, and the increase in the current is fairly small.

The 2' model (H562D) has the nominal current of 480mA with 22.5V Vf. So it is basically 2x H282D (1') connected in parallel. If you want to have 1 driver per shelf, you can go with APC-25-500 and 2x 2' strip serially connected. LPC-60-1050 can do 2 shelves (2 serial x 2 parallel). LPC-100-500 can handle 4 shelves (8 serial). LPF-90D-24 can be good with 7-8 all in parallel, and it is dimmable.

24W T5HO is about 2000 lumen, and 2' H562 (2') is about the same. But from my measurement, 2x H562 probably gives the similar (slightly less) PPFD. When I measured 4x 24W T5HO (with purple, "plant" bulbs), I got 160 micromol/m^2/s at 1' from the center. With my 4x 1' (H282D), I got 155 micromol/m^2/s. It seems to be weird that 8000 lumen of T5HO is similar to 4000 lumen of H series. This is partly coming from reflector loss in fluorescent light, and the output of fluorescent light decays after a couple months (mine was about 3 months old).

I'm not sure about the thermal glue. I got the one which was recommended by one of the rollitup.org member. With H-series, it has high efficiency, so heat management might not be too critical. Indeed, even without any heatsink, it wasn't becoming too hot. So you can give it a try.

If your electricity isn't too high, Samsung In-flux series is cheaper. Lower efficiency, but higher output with lower cost. But you might need to use a bit better heatsink than the C-channel.
 
  • #90
Bonjour

What do you think of these materials :

SmartArray Q36 LED-Module, 39W 4000°k 4247 lm and 2700°k 3700 lm or SmartArray L25 LED-Module, 32W 4000°k 3283lm and 2700°k 2865lm

LED Seoul Z P4, blanc, 240lm 6300°k 3.7w

The lumens do not seem to add up, is this true ?

jeff
 
Last edited:
  • #91
The 2' model (H562D) has the nominal current of 480mA with 22.5V Vf. So it is basically 2x H282D (1') connected in parallel. If you want to have 1 driver per shelf, you can go with APC-25-500 and 2x 2' strip serially connected. LPC-60-1050 can do 2 shelves (2 serial x 2 parallel). LPC-100-500 can handle 4 shelves (8 serial). LPF-90D-24 can be good with 7-8 all in parallel, and it is dimmable.

24W T5HO is about 2000 lumen, and 2' H562 (2') is about the same. But from my measurement, 2x H562 probably gives the similar (slightly less) PPFD. When I measured 4x 24W T5HO (with purple, "plant" bulbs), I got 160 micromol/m^2/s at 1' from the center. With my 4x 1' (H282D), I got 155 micromol/m^2/s. It seems to be weird that 8000 lumen of T5HO is similar to 4000 lumen of H series. This is partly coming from reflector loss in fluorescent light, and the output of fluorescent light decays after a couple months (mine was about 3 months old).
I wouldn't worry with highland Nepenthes, but lowland species tend to grow in shadier areas and be less light tolerant. I'm quite sure I would burn up all my seedlings putting them under light intense enough to grow sundews. Maybe if I run Nepenthes shelves at half power it will work.

Are you sure those purple bulbs are actually rated at 2000 lumens? I doubt they are. Fluorescent bulbs are less efficient at emitting longer wavelengths, and with bulbs so heavily weighted towards red that should lower the efficiency quite a bit. It's actually true of LEDs too, but they have higher phosphor efficiency in general due to converting blue light to longer wavelengths rather than UV, so there is less energy loss in the process.

I will certainly take a look at your recommended drivers. I wonder if perhaps I could set up 1 low light shelf + 1 high light shelf for now on a single driver? Basically 2x2' strips at full power and 2x2' strips at half power (Edit: I just realized that's what you meant by the LPC-60-1050 recommendation, but can that actually handle the voltage? I feel like the LPC-100-1050 would be required.) It seems like the larger dimmable driver would be a good bet and allow for easy future expansion, but the dimming system does not look like it's easy to use. Also do you have any suggestions for a strip arrangement for my 3' shelf? I was thinking I might need to do either 6x1' or 2x2' + 2x1' to provide even lighting for the whole shelf.
 
Last edited:
  • #92
Ok I did a bunch more reading and calculations. I reailzed that I would need a driver capable of 67.5V and 500mA (2 strips in parallel, with that in series with 2 more) to power two shelves as I mentioned, BUT I am calculating an average of 12,739 lux with 2x2' strips on a shelf. That seems a bit low for high light plants, though it may actually be perfect for lowland Nepenthes. It's making me wonder if 2 full power strips for a low light shelf and 4 for a high light shelf would be a better idea. But I suppose I should ask: How are your pygmy sundews doing in the tub?
 
  • #93
2000 lumen is just a ball park figure, and I'm not sure what my purple Odyssea plant bulb was rated at. I compared it to AgroBrite 6400K. At 1', a single bulb gives 475fc for Odyssea, and 472fc for the 6400K. So they are similar (Odyssea does use a little more electricity). But Odyssea gives quite a lot more PAR; 81 vs 58 micromol/m^2/s. It does seem to decay quickly over time, though.

LPC-60-1050 can provide up to 48V with 1050mA. Each 2' strip requires Vf of 22.5V at 480mA. So 2 strips connected in parallel will require a little more than 22.5V (each will get 1050mA/2 = 525mA). So if you connect two parallel circuits in serial, you'll need 45+V. So LPC-60-1050 works well for 4x 2' strips.

OK, if you want to 2x 2' for low light shelf and 3x 2' for high light shelf, ELN-60-24 (2.5A, 12-24V, so each will get 500mA, about $25) can do it in parallel. Or if you want to make it dimmable, LPF-60D-24 or PLC-60-24 has the similar spec for $30. I'm not so familiar with the following, but PLC-60-24, PLN-60-24, HLN-60H-24 have similar spec with dimmability.

For the shelf with 3' width, if you want to use 6' total of the strip, I think that it is easier to do with 6x 1' in terms of driver selection. APC-35-500 (or APC-35-700) can drive 6x 1'.

Going with a large dimmable driver could work, but you should note that if you dim the driver too much, AC/DC conversion efficiency started to suffer. So I generally try to much the driver and LEDs. For the flexibility, I recently learned an interesting driver LCM-60U (from this thread). With dip switch, it can change the driving current/maximum voltage combination, so this driver can be used for many different types of LEDs.

The dimming circuit is super easy. You just get a cheap 100k ohm linear taper potentiometer (cheap from ebay), and connect it to the driver. There are some variations, so I would buy a 5 pack or something, and use the multimeter to pick one whose maximum resistance is > 100k.

I think your calculation is right. So it might be better to go with 6x 1' strip at around 500mA each. Then you'll be getting about 6000 lumen. Your space is 3.25 sqft (36"x13"), so you'll get about 1846fc (=6000/3.25, about 20000lux) on average. In reality, it will be less since some light will be lost. To get this much of light with T5HO, I think you have to place the plants fairly close (slightly less than 1' from a 4 bulb unit). The continuous light of this intensity (with 12h day or so) is similar to 2800fc of noon intensity in the greenhouse (just a rough calculation). Drosera should be able to take more than this. My pygmy Drosera seem to be doing fairly well (but I might not know how they are supposed to grow) even though it doesn't seem to be a lot. Before this, they were getting much lower light. I should check PPFD when I get a chance.
 
  • #94
Bonjour

What do you think of these materials :

SmartArray Q36 LED-Module, 39W 4000°k 4247 lm and 2700°k 3700 lm or SmartArray L25 LED-Module, 32W 4000°k 3283lm and 2700°k 2865lm

LED Seoul Z P4, blanc, 240lm 6300°k 3.7w

The lumens do not seem to add up, is this true ?

jeff

Jeff, I'm not familiar with this unit, so I quickly looked it up. It is showing only 108lm/W, which seems to be a bit too low efficiency from the current standard. Also the price seems to be rather high. I think generation 7 Bridgelux Vero 18 or V22 should easily beat this module in most aspects (efficiency, total output, and price).

Are you wondering that 36 chips * 240 lm doesn't match with 4247 lumen? That's probably because the driving current is different; I'm guessing 240lm per LED is the maximum output, but the 4247lm is coming from a lower nominal current.
 
  • #96
Thank you for all the information, Naoki. I actually did run across the LCM selectable current drivers yesterday - I went through Meanwell's entire catalog trying to find higher voltage drivers that could handle 3+ strips in series. I really wish they had a product finder where you could input your desired voltage and current and it would tell you which drivers are compatible. Most models max out at <60V which isn't enough. I'm going to try to finalize my plans today then check Arrow tomorrow since they have a shipment of 2' strips coming in, though those may all be purchased already or have a high minimum order and force me back to Digikey's sky high prices.

on this module SmartArray Q36 LED-Module
SmartArray Q36 LED-Module, 39W | the leading LED-shop by LUMITRONIX
you have 36 x Nichia 757 LED

for 4000°k you have 4247 lumens
for 2700°k 3700 lumens

these lumens quantity is not suffisant for the plants ?

if I couple 2 led seoul (each 240 lumens) I have 480 lumens or always 240 ?

Jeff, the lights are bright enough for, say, 2 square feet (~0.1 square meters) but the LEDs have low efficiency so they will use more power and require larger heatsinks.
 
  • #97
Nimbulan, I went to check PPFD from the 4x 1' (H282D) on the pygmy Drosera. The leaves are about 9-10" from the LED, and in the middle, it was getting about 200 micromol/m^2/s. Around the edge, it drops down to 150-160.

Many CP people don't seem to talk about the light in a quantitative way. So at one point, I searched for scientific papers. I didn't spend lots of time looking for info about photosynthetic rates of carnivorous plants, but there is one paper by Bruzzese et al. (2010).

R. Bowler, H.B. Massicotte, and A.L. Fredeen. 2010. Photosynthetic light response in three carnivorous plant species: Drosera rotundifolia, D. capensis and Sarracenia leucophylla Photosynthetica 48(1):103-109 (I think you can get it from this link)

They used, GH grown D. capensis, D. rotundifolia, and S. leucophylla. As you increase the light intensity, phtosynthetic rate increase, but at one point, increase in the light doesn't increase the PS rate. This is the saturation point. Surprisingly, the photosynthetic rate (measured with a CO2 gas analyzer) saturated around 100-200micromol/m^2/s. It is surprising because this is the level close to the shade (under-story) plants. Field collected D. rotundifolia showed higher saturation point (400-500micromol/m^2/s). So they seem to be able to adjust (acclimate) their photosynthesis to different light environment. There is also something called a compensation point. This is the minimum light intensity above which plants can have a net gain of carbon. Below this point respiration rate is higher than photosynthesis rate. For the GH plant, the compensation points were 51, 78, 44 micromol/m^2/s for D. capensis, D. rotundifolia, and S. leucophylla, respectively. So 100-200 micromol/m^2/s isn't too bad for Drosera at least.

Jeff, here is a comparison:

SmartArray Q36, 25 euro, 108lm/W (I'm not sure what temp they used), 4247 lumen from 1050mA x 37.5V (=39.4W)
Vero18-C, about 13 euro, 156lm/W (@Tc=50C), 5583 lumen from 1050mA x 34.1V (=35.8W)

1/2 price, 44% more efficiency, 31% more total output (and it is easier to find a driver). So there is no reason to go with Q36. As Nimbulan said, it will grow plants for sure. But it is a rather expensive option in a short and long term. Here is the link to an example of Vero18-C: BXRC-40E4000-C-73 Bridgelux | Optoelectronics | DigiKey. If you want to get more output or higher efficiency, Vero 29 and V22 are the other options.
 
Last edited:
  • #99
Jeff, that Vero 29 is an older version. If you want 5000K, CRI70, this is the current Gen. 7:
BXRC-50C10K1-D-74 Bridgelux | Optoelectronics | DigiKey
In Gen. 7, Vero 29 comes in several variants. The one above is "36V" version (type D). There is also B-version (about 50V), and C-version (about 70V).

This is the link to B-version.
BXRC-50C10K1-B-74 Bridgelux | Optoelectronics | DigiKey

I think Vero 29 is the current best among commonly available and affordable ones. I haven't looked around recently, but with Vero 18, you can find heatsinks with predrilled holes, but you might need to drill holes by yourself with Vero 29.
 
Last edited:
  • #100
I suppose I still don't quite understand all the different units used to measure light intensity. Lux is easy to understand, at the very least, if not the most accurate way of doing it. What I've read recently is that most plants achieve chloroplast saturation at about 70,000 lux (though I'm not sure about the light source for that number) and I'm having a difficult time figuring out how that converts.

In any case, I find it odd that the saturation point would actually be different in greenhouse-grown plants. Would it increase if they were placed outdoors? Would the plants be just as healthy under each set of conditions?

I'm going to try not to overthink things and keep it simple. The tub you set up with these LED strips is a little over 2 square feet and 4' of LED strips produce 4000 lumens. When I first started I bought a 2x2' T5 fixture which also produces 4000 lumens (when new) for a ~2 square foot area (it's actually 3 square feet but the ends don't get enough light since the fixture's a bit short and I can't mount it any higher) and it has worked quite well. So I will stick with those numbers - try to provide 2000 lumens per square foot of grow space. That's actually what I aimed for when building COBs as well though the light from those isn't nearly as uniform.

For my old shelf that will be 6' of strips (2x2' + 2x1' or 6x1' due to the shelf dimensions) and for the new grow tent at 3.375 square feet per shelf I'll want 7 feet of strips but that won't fit evenly so I'd have to do either 6 or 8. Doing 6 (3x2') could be problematic since there is a support bar in the middle of each shelf so I think 8 (4x2') is likely the best way to go. I will be sticking with 2x2' strips for the low light shelf. I can always under-drive the strips slightly if I can find the right driver.

I'm hoping to get everything ordered today and with any luck I'll be able to assemble these lights by next weekend.
 
Last edited:
Back
Top