hoppycalif

The question about what happens to the light striking the front glass intrigued me. It is a simple first year physics problem, but it has been lots of years since I had that course.

Anyway I found, by googling, http://www.ps.missouri.edu/rickspage...efraction.html, which gives what we need to figure this out, and has a really neat graphical calculator that illustrates the theory. So, using that I came up with:



This tells me that any light that strikes the inner surface of the glass from the water, will be fully reflected if it strikes the glass at an angle less than about 41 degrees. Bigger angles let light escape out the front glass. Since the bottom part of the tank is further from the light, the angle will be the least down there. So, I ran to my aquarium (ok, I climbed the stairs slowly) and used a magazine to block the light escaping various parts of the front glass, by looking at the shadows cast by it on the floor. I found that for my aquarium, virtually all of the escaping light comes from the top third of the front glass, and none I can detect comes from the bottom third. I'm not yet sure how useful it is knowing this - one thing it suggests is that a light near the front of the tank, could reduce the escaping light from the front glass, and near the back could reduce the escaping light from the back glass. But, that back light would lose more light from the front glass and the front light would lose more light out the back glass. Yes, I am just brainstorming now. I may have made a mistake with those angles, but it is easy to get them from that graphical calculator.

newbie314

So if one raises the light higher off the surface of the water, the more of the light that get reflected and not escape though the glass.
Of course there is light instensity lost but I guess one could counteract that by a side reflector.

There is some inherent reflection of the glass as we all know when seeing the light bounce from our glass covers.

trag

That's exactly what made me think about this and decide it was worth scribbling some equations. As I mentioned, I took an electromagnetics class about a million years ago, and I vaguely remember those characteristics of point charges vs. line charges.

hoppycalif

Yes, the higher the light is above the tank, the smaller the angle of the light "rays" striking the glass, so a lesser amount escapes through the glass. But, of course, the higher the light is above the tank, the more light goes around the tank and is lost anyway, because of our very imperfect reflectors. The compromise would be an enclosure housing the lights, that has extra reflectors on the front and back of the enclosure, with a deep enclosure which moves the lights further from the tank. I stumbled on this design when I made my current fixture: http://www.aquaticplantcentral.com/f...t-fixture.html

hoppycalif

I have a 10 gallon tank, with two GE 6500K 20 watt powersaver screw-in CFL bulbs, each with a good reflector, but it is obvious that I only have a low light intensity in the tank. This has been a major annoyance to me. But, I think I understand it better now.

Those screw-in CFL bulbs are a somewhat spherical source of light - far from a linear source. So, their light intensity will drop proportional to the inverse square rule. My lights are one inch above the water, and 10 inches above the substrate. So, the light reaching the substrate is reduced by a factor of 100 from what it is at the top of the tank! If I had a 20 watt T5 bulb it would be a linear source, so the intensity would drop only by a factor of 10 at the bottom, giving me 10 times as much light as I now get. (It starts out being twice as good.)

I now think that is the major problem with the screw-in bulbs, much more so than the restrike problem. (I doubt that I lose a lot through restrike because I can look up at the bulb/reflector and see almost every surface of the coiled tube.) I am very tempted to retrofit the Perfecto hood with a linear bulb and reflector. It came originally with a 15 watt linear bulb that looks like a T4 bulb. So, that would give me another data point, an interesting DIY project, and might end the frustration with that tank.

BryceM

Before we all get carried away in a wave of point-source disgust and linear source passion, let's setp back for a second. You must be very careful to understand the assumptions that Trag is making in his post above. One must remember that modeling a CF or T5 as a true linear light source and a MH or spiral CF as a perfect point light source is a bit of a stretch. I understand the math and physics well enough. The equations are certainly valid. Their applicability to what we do doesn't pass the "real-world" test. Under certain conditions, with certain assumptions, fluorescents might out-perform MH's at a certain depth, due solely to their geometry. Assume, assume, assume ......... we're almost always better off to sticking with emperic data. Use what works for everyone else.

A thought experiment:

It's not much of a stretch to state that a watt of MH and a watt of T5 produce roughly equivalent quantities of useful light. Each is about 25-40% efficient at converting electricity to light. For our purposes, assume identical efficiency. Now, assume for a second that a "linear" CF or T5 fixture is truly 10X as effective as a "point-source" MH fixture for illuminating a remote surface. I know from experience that 3x150W MH HQI's are sufficiently bright to grow HC in a standard 180g tank with a substrate 24" away from the light source.

It's pretty clear that 45W (1/10th) of T5 lighting stretched along my 6' tank wouldn't do it. I doubt 250W of T5 would do it. In fact, I'd guess you'd need pretty close to 450W to get the job done. Where is the error? Didn't we all just accept the "fact" that linear sources are somehow an order of magnitude better? They should be at least 100% better, shouldn't they?

Well, the math for a linear light source assumes an infinite source length and a comparatively short distance to target. A 24" long bulb over a 24"x24"x24" portion of my aquarium still acts a lot like a point source. I'll also argue that my 3 MH fixtures spaced 2 feet apart over a 6' long tank act collectively a lot like a linear source.

Hoppy, your 10g tank is a low light tank for two reasons. First, spiral CF's are lousy designs for aquarium use. Even the best reflector is fighting the complex 3-D geometry of the bulb. There is a ton of restrike. There is tons of re-re-re-re-strike too. Second, the wpg rule doesn't work well for tanks that small. I use 8x23W spiral CF's over my 46g tank. It acts like a rock-solid medium-light tank despite the "apparent" 4 wpg.

I would absolutely love the chance to play with a PAR meter. On my 180g tank I have a combianation of MH's and T5's. It would only take a few mintues to produce loads of data with various lights on and off. It's hard to argue with data. Sadly, this is precisely what we lack during this sort of discussion.

hoppycalif

I am borrowing Tom Barr's PAR meter soon, so I will have some more data in a few days. I certainly agree that these simple equations don't accurately describe the non-simple light sources we use, but they do help to understand some of the characteristics we see. I doubt that any real light source has an inverse square drop off - maybe closer to dropping off at the 1.8 power of the distance. And real tubular lights probably drop off at something like the 1.3 power of the distance. I hope to get some usable data from both my AH Supply 110 watts of light and my 2X20 watt screw-in lights. I expect to be surprised again - I just don't know what the surprise will be. But, theoretical simplified considerations can help understand whatever the surprises are.

orion2001

Great points guaiac_boy and Hoppy. I'd love to see your data from your experiments with the PAR meter Hoppy. I agree with both of you'll in that the theory ofcourse is overly simplified but it can definitely help in understanding trends we might observe from experimental data. I'm also very curious to see your PAR data for your 10g tank with the spiral CFLs. Considering I have 2x15W CFLs in my 10g (emersed setup), I'd love to find out if I really need to boost the lighting up for my non co2, excel tank once I submerge the setup.

BryceM

Please experiment tons. I'd love to see the data.

trag

Another assumption/factor that I think is getting lost in the discussion is that for equal light production (roughly wattage) the linear source starts out less intense because the same light is spread over a larger area. The T5 intensity doesn't cross over to the MH intensity until a fairly substantial distance is reached.

Also, to emphasize guaiac_boy's point about the errors in treating a finite linear source as an infinite linear source...

Consider the target of a 4 ft. linear source with a 45 degree spread of light from the perfectly vertical. The light tends to act linearly, because each point on the target receives light from multiple point sources. The multiple point sources make up the linear light source.

However, the points of the target which are even with the ends of the 4ft. tube only get 1/2 the light intensity that the equations would predict. This is because they get light from multiple point sources on one side, but no light contribution from the other side. 1/2 of their light sources are missing as compared to an infinite linear source.

This effect is true to a lesser degree as you move inward from the end of the tube toward the middle.

Let's call the distance to the target, L. The point at which the light on the target is the same as the theoretical prediction is L inwards from the end of the tube. So, for a 4' bulb, if L = 2', then only at a theoretical dimensionless point at the very center, is the light as bright as infinite linear source theory would predict. If L = 18", then there is a 1' long area in the middle of the target where the light is as bright as theory would predict for an infinite linear source.

If the angle of light spread is 60 degrees, this effect is much worse. If the angle of spread is 30 degrees it is much better.

So, as G_B says, don't get too carried away. Linear bulbs have some advantage over point sources in light spread--certainly more than some MH proponents would allow. But they are not infinite linear sources, unfortunately.

I too look forward to seeing more PAR meter readings. Especially interesting would be reading from end to middle on fluorescent tubes at varying depths.