I mentioned in another post about an efficient T8-LED which I recently found, and I think someone wanted to know about it more (it could have been in a different CP forum).
I put some of my first impression and data to my (brand-new) blog:
I hope that it may be useful to some people.
Chief Cat Behavior Specialist
Very descriptive in your blog naoki. I have similar T8s w/out the led tubes used for random plants. Thank you for sharing & for all of the pics.
Am I reading this correctly? Two of those T8 bulbs replace 1 T5HO in terms of light output and will use about half as much energy?
Originally Posted by naoki
And, I didn't see anything about whether light output stayed constant. Have you had them long enough to tell? (I know the T5s lose output over time, and, I thought, a similar thing was said to happen with some of the original LEDs, but I'm not sure).
Natch, you are correct. But that estimate is from a point PPFD measurement, so it is somewhat a rough number. The caveat is that the beam pattern (spread of light) influences the interpretation. In other words, if the LED had a really narrow beam angle, the point measurement would tell me that the LED is much more efficient than the real efficiency advantage because all of the light would be concentrated on the place of the measurement. But I think that the estimate may be semi-decent from the following reasoning.
Originally Posted by NatchGreyes
From the factory specs, Sunritek T8 LED has 60% higher luminous efficacy than T5HO. From my measurements, if you calculate the footcandles/W and compare T8-LED and T5HO, you'll get (485/25.8) / (590/50) = 1.59. So this is similar to the efficacy ratio based on the factory spec (total lumen/W). This suggests that the influence of the beam pattern may be somewhat small in this particular case. So in terms of PPFD, the 100% higher efficiency (i.e. twice as efficient as T5HO) could be a good approximation. You may be wondering why there are only 60% efficiency advantage in terms of human eye brightness (lumen or foot candle), but there are 100% advantage in terms of PAR. This is basically because the emission spectra of florescent lights are very different from that of white LEDs (even if you compare 4000K vs 4000K). So the amount of PAR for a given lumen is different.
I think that you are talking about the bad lumen maintenance of early LED strip light (the ones you cut into a length by yourself). That was caused by a bad heat management (driven too hard/hot, and heat was killing the LED performance). Basically the majority of strip LED light isn't designed to be efficient. Similarly, it could be an issue with lots of cheap LED fixtures, which are pushing the driving current to save the cost of manufacturing. So for example, you can make a 100W LED fixture by using 20 x LEDs which is running at 5W each. Or you could run each LED soft (let's say 2.5W for each) and use 40x LEDs (exactly same diode), which gives the total of 100W fixture. The later option is much more expensive initially, but it will increase the efficiency quite a bit. This is the reason that there is generally a trade-off in the cost and efficiency of LED fixtures. Then if the manufacturers of cheap LED fixtures cut the cost by using insufficient heatsinks, then the longevity (and possibly lumen maintenance) can be influenced.
Sunritek product seems to be using fairly soft-driven LEDs. This means that the longevity should be better. It is true that the heatsink doesn't look so impressive, but it doesn't heat up at all (a good thing). But we don't know how the longevity and lumen maintenance of this product will be (it is too new).
Last edited by naoki; 08-16-2015 at 03:16 AM.