What type of T5 bulbs do you use?

[naoki]

According to a couple web pages (google T5HO lumen maintenance), T5 (and probably T5HO) seems to have a longer life than T8.

http://ecmweb.com/archive/t5-fluorescent-lamp-coming-strong

With 13h/day, it is 4745h/year. With these figures, they are assuming continuous lighting. In reality where we turn them on and off once a day, the light bulbs wear faster. Still I think I can go for 2-3 years (actually I would probably move to all LED).

There are also the issues of color shift in addition to the intensity decay. But I'm not sure how it will influence the plant photosynthesis. Isn't it possible that the spectrum becomes better when warmer bulb shift toward more blue output? But I don't think we can find the PAR (instead of lumen) maintenance curve published anywhere.

 

[TyroneGenade]

From an aquarium perspective I have had a look into lamp lifespan and like Naoki, I haven't found the "6 month" rule to be supported by the data...

A friend and I have been comparing lamps for several years now based on plant physiology and lamp physics data and it seems that warm whites and daylights are about half as efficient compared to "plant gro" lamps like Gro-lux and Fluora-Glo when it comes to useable photons... I had originally rejected this idea as I had grown plants just fine under daylight and cool whites but the actual data speaks against this when the plant growing conditions are ideal.

So, with that said, if you are seeing such a big improvement with the plant-grow lamps then its a sign that you have otherwise near ideal plant growing conditions. We look forward to seeing photos of your flowers.

Incidentally, we find almost no T5 advantage over T8 tubes (http://www.apsa.co.za/board/index.php?topic=4454.0 ) as they both employ the same physics and chemistry. Where T5s have a big advantage is in their small size (relative to T8s) that enable you you squeeze more watts of light into the same space and that their narrowness means that less light is lost when you employ a reflector. A good deal of light gets lost in a reflector when you use a thick T8 or T12 tube. Geometry matters (http://www.apsa.co.za/board/index.php?topic=4379.msg41393#msg41393 ).

I still vote for LEDs over tubes. They are leaps and bounds way more efficient than tubes. Watt for watt, you may need only a 5th of T8 watts if you run LEDs instead.

 

[naoki]

Great info, Tyrone! I have to study the german link you posted there (there are so much interesting info about plants in German..., I should have studied German harder when I was an undergrad).

The calculation of efficiency posted by Greystoke in the first link is interesting. I didn't know that it is so high for fluorescent light (about 27.5% of electric energy is converted to visible light energy). I think that with modern white LED (e.g. Cree CXA-3070) and 90% efficiency constant current driver, you get about 27-36% (higher efficiency with lower driving current). But CXA-3070 is one of the top end (for commonly available LEDs with reasonable cost) in terms of efficiency at this moment. I finished drilling/tapping the heatsink, so I hope I have some time to test out my CXA-3070 rig this weekend!

You are right that the LED is directional (an advantage over T8), but "LED watt = 5th of T8 watt" seems a bit too extreme, doesn't it? With white ones, I have thought that you can go by reducing 10-20% of watt to get similar PAR (i.e. 25-29W LED instead of 32W T8).

 

[TyroneGenade]

You are right that the LED is directional (an advantage over T8), but "LED watt = 5th of T8 watt" seems a bit too extreme, doesn't it? With white ones, I have thought that you can go by reducing 10-20% of watt to get similar PAR (i.e. 25-29W LED instead of 32W T8).

At first we were predicting somewhere between 2 and 3 times more potent but then I found this: http://www.advancedaquarist.com/2013/3/review and http://www.buildmyled.com/12000k-reef-spectrum/ (see produce specs lower down). That 48 inch LED is about 100 W/m2 and is getting about 110 PPFD at about 17 inches water depth. You would need about 300 W/m2 of T8/T5 lighting to do that and my math is VERY generous with respect to reflector efficiency and how much light actually enters the water (my model is buried in that first APSA thread). Somewhere around 3x more effective looks like the best guess (I struggle to recall my reasoning for 5x... maybe it was a typo?).

I think we could safely cut out power consumption by a third if we switched to LEDs for growing.

PS. most of the above only really applies to fishtanks. Even with a really good reflector only about 75% of the light is focused onto where it is supposed to go so LEDs (which focus light downward) are at least 1.33x as efficient as T5/8 lamps. If you are not using reflectors then only about 25% of the light is reaching your plants, in which case LEDs are 4x as efficient. Assuming you want to mimic shade (20000 lux) or overcast light (2000 lux), this would amount to between somewhere between 190--250 W/m2 or 19--25 W/m2 of LED light if you chose the correct beam angle to have all the light focused on the plants. The plants directly under the lights would have 2.5x the lux as those on the periphery so you can distribute the plants under the lights based on what you think they really need. For comparison, you would need 433 and 43 W/m2 with T5/T8 tubes (if you have a great reflector!) so theoretically, LEDs are 2.3x more efficient for our purposes. Where the Lux-based math breaks down is that plant-gro lamps typically have a lot less lux (but still pack a big photosynthetic punch). Typically, they have about half as much... so it is better work on the wattage than the lux. You need half the wattage of plant-gro lamps per watt of cool white/daylight lamp.

 

[silence882]

Thanks for all the info!

I think I'll try this set for a year and see how the plants do. If they still look good I'll wait until 2 years have passed to replace the bulbs.

I had another bad experience with only using one type of bulb many years ago. I think I am now a believer in using at least 2 types in each fixture.

--Stephen

 

[silence882]

The new GE bulbs arrived! I now have quite an array in my fixture.

From left to right: GE Starcoat 6500K, Aquamedic Aquiline, GE Starcoat 3000K


--Stephen

 

[naoki]

Stephen, they look quite different in photos (human eyes seem to minimize the difference).

Tyrone, here is some data comparing the efficiency between LED vs florescent light:
http://cpl.usu.edu/files/publications/factsheet/pub__7199661.pdf
Table 3 Photon efficiency column. Super-high-end LED can be 2x more efficient, but most of LED (consumer-grade high-end) is only 19% more efficient than T8. I believe that efficiency of T5HO is similar (or slightly lower than) T8.

 

[polyantha]

Stephen, they look quite different in photos (human eyes seem to minimize the difference).

I noticed the same. My fixture with daylight and blue light was extremely white and blue on photo and even on the camera screen.

 

[TyroneGenade]

Thanks for the data Naoki. But, just to be clear, the principal advantage from LEDs is the optics that produces a narrow beam of light. Even with a good reflector (most reflectors are not good as the tubes aren't placed deep enough into them) at least 25% of the light from a tube is wasted, and then the rest spreads out and grows "dilute" very quickly. While LEDs are only 19% more efficient at turning electricity into light, the optics makes them 1.3--2x better at actually getting the light to the plants.

 

[Ray]

From an aquarium perspective I have had a look into lamp lifespan and like Naoki, I haven't found the "6 month" rule to be supported by the data...

The folks at Hydrofarm tell me that one year is probably the safest/most economical point, while still ensuring decent lighting. When I have run a T5HO for 12-14 hours per day for a year, then replace it with a new one, I DO notice a difference in the brightness.

Shifting gears a bit - I now carry the Philips Green Power LED lamps (so far I only stock the White/Deep Red Flowering bulbs that screw into an incandescent socket), and am thinking about buying a couple of the Deep Red/Blue production modules (35W, 120 mm length) for my own use. They normally retail for as much as $240 each, but if I buy a case of 6, I can get them for about $150. Anyone want to jump in on that case?

 

[Linus_Cello]

The folks at Hydrofarm tell me that one year is probably the safest/most economical point, while still ensuring decent lighting. When I have run a T5HO for 12-14 hours per day for a year, then replace it with a new one, I DO notice a difference in the brightness.

Shifting gears a bit - I now carry the Philips Green Power LED lamps (so far I only stock the White/Deep Red Flowering bulbs that screw into an incandescent socket), and am thinking about buying a couple of the Deep Red/Blue production modules (35W, 120 mm length) for my own use. They normally retail for as much as $240 each, but if I buy a case of 6, I can get them for about $150. Anyone want to jump in on that case?

What does this "screw/insert" in to? The link mentioned "conventional flourescent." So like a T12?

 

[Ray]

No, the ones I have in stock screw into a standard incandescent, "Edison Base" socket:

The "Production Modules" I'm thinking about are linear, stand alone lamps. Here's a video on someone experimenting with them for lettuce growing:
http://www.youtube.com/watch?v=u8-FKHAVsnA&list=UUkhj6cskGZ3fOU0uGyoqnoA&feature=share

 

[naoki]

Thanks for the data Naoki. But, just to be clear, the principal advantage from LEDs is the optics that produces a narrow beam of light. Even with a good reflector (most reflectors are not good as the tubes aren't placed deep enough into them) at least 25% of the light from a tube is wasted, and then the rest spreads out and grows "dilute" very quickly. While LEDs are only 19% more efficient at turning electricity into light, the optics makes them 1.3--2x better at actually getting the light to the plants.

Tyrone, take a look at the Methods section. The efficiency they measured is at the fixture level, so it is not comparing just the diodes. The data incorporate the efficiency of reflectors/lens and ballasts/controllers/drivers. They didn't mention a specific model of florescent fixture or bulb, though.

Ray, that's cool you are carrying these Philips GreenPower LED. They seem to have really high efficiency (if the numbers in the specification is correct). For the linear 4' models, the photon efficiencies are 1.43 and 1.67 micro-mol/J for Red/Blue and White/Red, respectively. These numbers are comparable to the best LED/HPS from the Utah study. It is interesting that White/Red has the higher efficiency than Red/Blue. The screw-in type is pretty efficient (1.22 micro mol/J), too. The linear one would be a nice replacement for florescent bulbs. They are surely pricey initially (even with your volume discount), but you probably save money over the cheaper LEDs in a long run. IP66 means that they are water-proof, right? They look like that they are fan-less, passive cooling, which is great for indoor, too.

I wonder what kind of White:Red ratio, philips GreenPower is using (I didn't see the info in Phillips site). I did make the public lab DIY spectrophotometer, and it works, but it wasn't good for my camera. Now I made a second version with a Cheerio box, which is much easier and seems to have a better resolution. I still need to figure out how to take photos with it. Maybe you can try it and measure the spectra.

 

[Ray]

IP66 - the first 6 indicated dust proof, the second is the water intrusion level, 6 being protected against strong water jets.

They don't tell you the ratio, but here is the spectrum. The title at the top is reversed. That red peak reaches an absolute irradiance level of about 52 µW/cm2/nm, while the red on my little ones only reached 4. My blue peak is 13 compared to about 6 for the Philips lamp

 

[junglejim]

I have very good results with Rays LEDs over my paph seedlings as well. (distance of 12") Many of them are in bud for the first time.

Are you using the LED vivarium 13 watter, or the ones that replace the florescent tubes in light carts?