How creative are you?

Hi all,
I'm obviously new here. What brought me here was a need to design, from scratch, an LED system to light up a fiberglass tube from the inside. Cost IS absolutely a factor. However, if it won't do the job, it doesn't much matter what it costs, does it?
Parameters:
1.Total diameter not to exceed .900"
2. Run time on fresh cells= 10hrs or more
3. This is the only one that I'm having trouble being specific on. All I have so far for a reference is a Garrity Steel Beam AAA. It illuminates the tube, but barely. Not acceptable. The Garrity is bone stock. I've searched to no avail to find a spec on the Garrity for its candlepower or luminosity, so I couldn't compare a commercially made light to see if one that is mass-produced would work.
So, what I'm asking for is a parts list or schematic for a cost efficient LED circuit that is VERY bright white light. I've read on this board about heat becoming a factor at some point. The system will be housed inside PVC then installed in the tube, so airflow for cooling will be nill.
If any of you are familiar with it, what I'm constructing is similar to the way a borescope works, except this one will be housed inside the tube it is illuminating.
End result if done correctly----fiberglass tube will glow "brightly" in the dark!
Any help you could provide would be greatly appreciated as all of you have much more experience in flashlights and their construction, than I do.

Well, there's laserbrite and glo-toob for a start. There are several others.

Not knowing what you mean by "very bright" is also a bit of a problem.

Mike Painter said:

Well, there's laserbrite

Click to expand...

Hard to take a company seriously when their demo video has at least 5 mispellings. (lazer, brite, emulate's, tuff, tanslucent)
Then again, it does look like a decent and innovative product. I'll be keeping my eyes out for it.

To answer the OP's question, sounds like you need the brightest 5mm LEDs you can get, or at least, not Luxeons.

Mike, I really don't believe the glo-toob will be bright enough. The fiberglass tube is not transparent, but somewhat opaque. I guess the proper term would be translucent. And as far as the laser-light, well I have to agree with carrot.

Carrot, yes, the 5mm LED sounds right. Since I am not nearly as familiar with LED's as are you guys/gals, I just don't know which one(s) to try first. It seems through surfing the web, that it is a "crap shoot" for me having little experience. I really have no grasp of the difference between 500mcd and 1200mcd.
I guess what I'm hoping to achieve here is an "educated guess" from some of the brilliant minds on this board. Obviously, from the parameters given in the OP, a C or D cell is out of the question. So, if limited to a AA or AAA cell, it makes the equation more difficult for me to answer. If no one were to have a suggestion that they thought were a good starting point, then I'll just try as many as necessary to get to the target. I'm just hoping to shorten my learning curve a bit.
Thanks.

Infosponge said:

Mike, I really don't believe the glo-toob will be bright enough. The fiberglass tube is not transparent, but somewhat opaque. I guess the proper term would be translucent.
Thanks.

Click to expand...

I mentioned the glo-toob as an example of what has been done, not as a source.

You mentioned a .9 inch diameter. This will eliminate C and D batteries.

Your best bet would be to get a piece, stick the end in a box with a big light and a dimmer to geet an idea of what you need.

I don't know about LEDs as much as some other people... but IIRC, larger mcd number is higher output. If you want really high output you'll have to jump to Luxeons, which do create a fair amount of heat... not something you really want unless you can plan out a way to get the heat out.

Nichia CS is a good 5mm LED. It's used in the Arc AAA-P (you can see the output ratings for that light here: http://flashlightreviews.com/reviews/arc_aaa-p.htm ). The output of an Arc AAA-P using a 5mm Nichia is rated at 5h on high and another 5h on low. This makes the Arc AAA-P an interesting candidate for you to consider using as a base, although runtimes aren't as high as you want. Throw in a lithium AAA and you'll get a slightly longer runtime as well as a slightly higher output. People say that the Nichia CS is more of a blue-white than a snow-white, so it may not be optimal for you, unless perfect color tint doesn't matter.

You might want to check out the Nichia website: http://www.nichia.com/product/index.html

I hope some real LED junkies wander by here soon to help me out, I'm almost out of my league here, though I've tried to present some useful information. Infosponge, you may want to draw attention to this thread from the LED board in General Flashlight Discussion.

You might also be interested in checking out this thread I posted here - you can try and order a few (free) LEDs to sample from some suppliers to see if they meet your needs. Also good for getting a few boards and circuits that may help you, although I don't know enough about the circuitry to help you there. https://www.candlepowerforums.com/threads/101337

Thanks guys! I really appreciate your help. I have some sample LED's on the way to test. I'll get back with the results as soon as possible.

Your requirements seem to preclude luxeons; they require heatsinking for either high-power operation or extended runtimes. Depending on how long this devise has to last, you'll probably end up with either 5mm LEDs or low-power surface-mount LEDs.

A word on candella or mcd (milli-candella) ratings - they are spot intensity numbers, typically for the center of the beam, and often exaggerated. Pop quiz : which is brighter - 10,000 mcd over 15 degrees or 400 mcd over 120 degrees? Surprisingly, 400mcd over 120 degrees is more than twice as much light as 10,000 mcd over 15 degrees (~1.3 lumens vs ~0.5 lumens).

You're going to have to scrutinize spec sheets very carefully. Most low-power white LEDs are specified at 20mA with a Vf (forward voltage) of 3.6V. Sometimes a manufacturer will tweak a stat or two away from the baseline (like specifying mcd at 30mA - which drasticly reduces life, increases heat production, and makes the device less efficient).

You could take a look at the LED DB link in my signature for some rough analysis of commonly-available LEDs ... bear in mind that the lumens approximations from beam angleXmcd are very rough, and are done purely to simplify comparison between similar items.

Others might be able to help you on electronics to drive this device. You cannot treat LEDs like filament bulbs - they need both sufficient voltage and current regulation.

What type/size of battery? Since cost is a factor, would rechargeable be an option?

If your tube doesnt need to be verry durrable you could try cold cathode tubes. These will run very bright, white, and efficient, and cheap: http://www.oznium.com/cathode-kit

how about using UV leds and white fluorescent paint on the tube.....