Ultra bright LED 1000 lumen plus ???

[Andrew Mawson]

I am looking to replace a microscope bulb in a piece of equipment with something somewhat brighter than the "13347W" 6 volt 15 watt BA15 based T6 shaped incandescent that is in at the moment that apparently gives 210 lumens. I bought a cheap torch claiming an output of 2200 lumens from a "CREE XM-L T6" LED intending to rob the LED module, however I suspect that the lumens claim is fictitious as it seems no brighter than the bulb.

Does anyone know of an LED module no more than 19 mm in diameter producing white light of 1000 lumens or more ? (operating voltage / current unimportant)

Also I'd like to find a data sheet for the "CREE XM-L T6" if anyone knows a source - the CREE web site isn't helping :(

Andrew

[Andrew Gabriel]

Not possible in that size - couldn't dissipate enough heat and stay within LED junction's operating temperature. That would be an LED with an equivalent mains filament lamp output of around 70W.

However, you don't need anything like that light level. A 3W LED will easily give >210 lumens, but as they can be directional and/or focused, the light intensity can be an order of magnitude or more higher than your 15W filament lamp.

[The Other Mike]

The first document details the binning that the manufacturer performs post manufacture to group the LED's by output which is the T6 section of the part number. The second is the datasheet, 900 lumens is about the limit when driving the device at around 3A, heatsinking could prove 'difficult'

Why the need for all that light when the exisiting installation is just 210 lumens?

Safety!

P.S. There are lot's of fakes about

[Adam Aglionby]

LED macro ring?

[Dave Plowman (News)]

You'd need to define 'white light'. Problem with LEDs is producing a reasonably consistent spectrum. Which comes for free with tungsten.

Of course this may not matter for your application.

[Andrew Mawson]

Thanks for those links, much appreciated.

The image is projected onto a 3" diameter ground glass screen - the application is setting tool height on a CNC lathe.

[Andrew Mawson]

Nice bit of lateral thinking - I'll experiment

Andrew

[Martin Brown]

Assuming therefore that you can machine and tap your own aluminium heatsink why not buy a bare batwing LED array in the 7W category and mount it yourself so that it fits where the old filament bulb went.

You might need a condenser lens or two to make the light parallel if it is intended for microscopy and you choose a single die device. LEDs aren't quite point sources but they are not all that far off either.

Getting the relatively small amount of heat away will be important. An LED doesn't like to be too warm or the junction degrades quickly.

[Andrew Gabriel]

I've used some of those. Seriously bright, so be careful during the assembly/testing (run from much lower current to avoid full output in your eyes).

BTW, it's a lambertian light distribution (which is what you want here), not batwing. (I don't think you can buy any batwing distribution LEDs anymore - they all vanished just as I wanted to buy some;-)

Also, this module is not very temperature-stable - it really does need a constant current supply, and not a simple series resistor.

[alan_m]

LEDs come with different beam angles and may require different lenses in front of them. A LED in a touch may be optimised for a wide beam angle and the cover glass on the torch is likely to be focussing the light. You may require a LED with a narrow beam. Try looking at the power led section of the

site (there is a switch for English web pages) and/or the optional optics for the LEDs. You may/will require some heat sinking. Excessive heat will result in a short life span.