The Future Has Arrived

[Stefano]

So if you just have UV leds would work to cut out blue leakage.

The problem is that UV LEDs, especially at short wavelengths (it seems) emit a yellowish-white light, at least up to red (and maybe even a bit of IR). Nichia's LEDs (like the ones used in Convoys) are known to emit very little amounts of "white" light, and the main problem is blue light. In that case a yellow filter should work quite well (even if I would use a black U-glass anyway to be sure of blocking everything), but I am 99.9% sure that Col's LEDs emit some yellowish-white.

[ulf]

I could mount the filter at say 300mm from the LEDs, with a tube attached & add a ventilation chimney with a 12v computer fan to pull out the heat....?

To cool the filters you only need a relatively thin passage of air. As you must lift the filter away a bit over the LEDs to avoid them touching use that airspace for cooling too.

10cm distance or slightly less would be OK as long as the fan-propelled air mainly pass by the filter.

 

I would direct the hot air exhaust backwards to avoid disturbing your photo objects. Rocks are immune, but some lighter objects like flowers will otherwise move around due to the airflow.

[colinbm]

Thanks guys,

I have tried a 90mm dia x 2mm wall PVC tube 50mm long with the filter at the end.

The PVC gets to about 60deg but not hot enough to deform, the filter was around 50deg, so with some forced ventilation, I'll try it again tomorrow.

Col

[dabateman]

Lee 010 (25%) and Lee 100 (40%) in the 310nm to 370nm region then shoot up over 460nm.

Lee 514 looks to have about 55% AT 310 to 400nm, then shouts up over 450nm. So may not work if you have a really leaky LED.

[ulf]

Lee 010 (25%) and Lee 100 (40%) in the 310nm to 370nm region then shoot up over 460nm.

Lee 514 looks to have about 55% AT 310 to 400nm, then shouts up over 450nm. So may not work if you have a really leaky LED.

Those foils will be rather hot fast without forced cooling.

Assume 100W input electrical power, conversion efficiency to optical energy 30% => 30W. 70W cooled away from the LEDs.

Lee 514 with the best transmission must absorb and dissipate 13.5W over a relatively small surface area, especially as the backside without forced cooling will be a static air-pocket to avoid light leakage around the filter's edge.

[dabateman]

But they are designed for that. When I did stage lighting, the hallogen lamps would just cook out crazy heat.

I think I only melted one gel once and it was stacked to get a deep blue. It was also in a crazy intense spot light.

[colinbm]

Further satisfactory progress....

With this cooling of the light front, it is stopping the heat from getting to the filter.

This is over kill & unappealing, but it is a prototype & proves the concept.....now to simplify....

I am waiting on another liquid CPU cooler & I will modify it so it can be mounted with the fan above the light front & add a chimney closer to the light front so the filter can be closer to the LEDs.

This is what I tested with today.

 

[ulf]

IMHO this multi-wavelength LED array is not especially well suited for filtering with traditional filters like UG1,UG11 or their Chinese copies.

Those filters absorb most of the energy from the light emitted from the LEDs with longer wavelengths.

 

Before my similar project stalled I measured the spectra from my three types of LEDs with peaks at 369nm, 386nm and 393nm.

Their combined spectra looks like this:

Please compare that with the transmission of UG1 and UG11 in approximately the same wavelength range:

 

It is evident how badly the peaks of the two graphs overlap.

 

Except for light from the 365nm LEDs almost all supplied energy will be converted to heat.

I think this is a complicated and expensive way to heat the room.

 

A better alternative would be to not try to combine illumination for fluorescence photography with UV reflective photography.

Maybe a second LED-array with only 365nm LEDs for a filtered output would be a better idea?

[colinbm]

Thanks Ulf

I only have a U340 at 77mm x 2mm to place in front of this LED, I am not sure what the final output would be.

I have asked if another set of UV LEDs could be made, but with out the 395nm & with one lower then 365nm in its place.

I enjoy the experiment & challenge....

[ulf]

The U-340 2mm is rather close to the UG11 2mm in the diagram above.

It transit light mainly from the 365nm LEDs. Most of the other light is lost.

 

Experimenting is a good thing to do to discover and learn.

Please do not see my comments as criticism, but as an intent to help and add my conclusions.

[Stefano]

The U-340 2mm is rather close to the UG11 2mm in the diagram above.

It transit light mainly from the 365nm LEDs. Most of the other light is lost.

That’s why I would use U-360 or UG1 (maybe thin, if they don’t heat up too much). They reduce longer wavelengths less.

[colinbm]

Thanks Ulf

Yes I perfectly understand, you weren't being critical, but I take criticism OK too.