Number of colored Cree XP-Es to match a 12V Cree MKR or another solution?

[Rimbaldo]

Hi!

I have adapted a dental led focusing light (ovehead lamp), that used to have an halogen lamp, to a LED one, using a 12V 15Watt Cree MK-R Neutral White Led. The dental overhead lamp has a kind of parabolic glass mirror that focus the light to the client's mouth.

But I want to add to this MOD I've done, an orange or yellow colored LED, so I can choose between the Neutral White and the orange/yellow one. I don't want warm white, I want orange color (or yellow).

Unfortunately there's no Cree MK-R in orange or yellow, so I though about using 4 Cree XPE-s (700mA each) in orange or yellow in series, around the MK-R, so I can use my same 12V supply that powers the MK-R. I'll never use both colors at the same time. It will be white or orange, never both altogether.

But I've tested the orange bead on the bench and it's really really faint compared to the MK-R. I'm suspecting that even adding 4 beads in series I won't have the same lumens but in orange (or yellow) that the MK-R provides in white.

So, could anyone help me? How could I have an equivalence of lumens of a Cree MK-R 12V 15W single Led, but using colored (orange or yellow) leds? It could be even be another type of Led also. I prefered some led that could be put around the MKR, so I can still have the light focused.

In my 22mm diameter Aluminum pcb I'm designing, I have the MK-R in the center and 4 XP-Es around it, but I suspect even 4 beads of orange won't do the job I want done.

I wanted to use the same 12V supply I'm using, if possible...

Thanks!

 

[LEDphile]

The MK-R is roughly equivalent to 4 XP-G2 LEDs in series, so if you are running XP-E2 LEDs at 700mA, you'll need ~8. PC-amber XP-E2 LEDs will be subjectively close in output to the white LEDs when driven at the same power level, as will red-orange (note that with red-orange, you will need more LEDs than with PC-amber to bring the string Vf in line with the MK-R). Direct amber isn't going to give you the output you are looking for.

If the optical system is the type I think it is, your LED array should be laid out in a linear fashion to mimic the halogen filament (this will also likely be better from a thermal perspective), although there is a good chance your array size will end up larger than the halogen filament. If you need to shrink the array, you might consider some of the smaller LED packages (e.g Cree XQ-E family, although there are others).

And finally, if you are using a constant voltage power supply directly to the LEDs, there's a decent chance that your drive current is not what you expect (and may be exceeding the specifications for the LEDs). This can also contribute to brightness discrepancies.

 

[Rimbaldo]

The MK-R is roughly equivalent to 4 XP-G2 LEDs in series, so if you are running XP-E2 LEDs at 700mA, you'll need ~8. PC-amber XP-E2 LEDs will be subjectively close in output to the white LEDs when driven at the same power level, as will red-orange (note that with red-orange, you will need more LEDs than with PC-amber to bring the string Vf in line with the MK-R). Direct amber isn't going to give you the output you are looking for.

If the optical system is the type I think it is, your LED array should be laid out in a linear fashion to mimic the halogen filament (this will also likely be better from a thermal perspective), although there is a good chance your array size will end up larger than the halogen filament. If you need to shrink the array, you might consider some of the smaller LED packages (e.g Cree XQ-E family, although there are others).

And finally, if you are using a constant voltage power supply directly to the LEDs, there's a decent chance that your drive current is not what you expect (and may be exceeding the specifications for the LEDs). This can also contribute to brightness discrepancies.

HI! Thanks for your answer...

It would be easier then to use 4 XP-G2, but they are not colored...

The halogen light filament was perpedicular to the curved mirror, so I think a circular array would work, but I can make it in a linear fashion also.

So if I plan to use the XP-E2, I would need 8 beads of yellow PC-amber) , or even more of the orange/red, because of the orange's lower Vf??

My constant current-led Driver outputs 900ma at max 12w, which supplies the MK-R.

So in order to supply 8 beads of the XP-E PC-amber I would need at least a 24W constant current driver to get the same "brightness" of the MK-R?

As the XP-E2 withstands 1A of max current, I could use two of my led drivers, each one driving 4 beads in series at 900ma each? Am I correct?

One more thing.. these 8 Pc-Amber leds, at 900mA (or even with a driver set at 700mA) would generate much more heat than a single MK-R, to allow the same "illumination"?

I could use another ones smaller also: XQ-E or XB-D. They are also available in PC-Amber

Would you recommend any of these?

Finally.. over digikey is this one the XP-E2 PC-amber? Digikey Link


Thanks a lot!

 

[LEDphile]

If you are running at 900mA, the only thing XP-G2 buys you over XP-E2 is slightly better efficiency. And you can likely get away with 4 PC-amber XP-E2, since you are under the 1A max current for the part (the recommendation for 8 assumed you were driving the MK-R at the full 1.25A, which is above the limit for the XP-E2). Moving to smaller packages (e.g. XQ-E) is fine, but you will take a small efficiency hit and thermal management becomes more challenging due to the lack of a dedicated thermal pad.

If the original halogen lamp used a compact filament instead of a linear filament, you may be running into optical source size challenges - the size of your LED array may be enough larger than the halogen filament that the light from the outer LEDs is not captured by the optical system.