Epileds 5050 (XM-footprint) high power red led, short post with test results.

I stumbled upon these a few months ago and did an output test on one, but because the performance is not that great the urge to make it public was not high. In another thread Flash Tom asked about it so here are the raw measurements at least.

The die is larger than the XP-E2 and the output colour looks exactly the same. Here is a picture next to an ancient red XP-E (same size as XP-E2), to get an idea of the size.

Here are the numbers, I re-tested a red XP-E2 next to it because it was two years ago that I tested it, and the results to my contentment are virtually the same.

So, as I announced, despite the considerably larger die (2x ?), this led performs just a littlebit better (from 2A on). I now see that the voltage is quite a bit lower than the XP-E2, that adds to the efficiency, I may have underestimated this led a bit. And they are not expensive and should make a nice 12-led SRK-clone red monster.

1 Thank

Thanks :beer:

Yep, Thanks also. Iā€™ve been using the red XPE2 for awhile now in B158ā€™s. Thanks to your tests I know what current to drive them at. :+1:

Thank you very much djozz! Iā€˜m already enthusiastic about trying out this LED in some hosts. :slight_smile:

Have you tried de-doming them? I know with the cheap Chinese knock-off emitters, chemical de-doming is sometimes super difficult, if not impossible. Maybe slicing?

There have been talk about and attempts at dedoming colour leds before, with no clear gains. Whatever is the plain optical benefit of removing the dome (like a possibly better behaviour in TIR optics), these dies do not have a phosfor layer that can re-excited by reflected blue light so the added luminance increase of photon recycling is not there.

I would have thought there would be some gain from a smaller apparent die size.

Thanks for the hard work mate!

The bigger LED size would have a better look in a floody reflector light?

I have not checked but the die size is still not big, as small as XP-G or 219C, should work in any reflector. The die not square, I wonder if that has any influence on the beam (my guess: not)

Since you said it is as small as XP-G or Nichia 219Cā€¦would this LED fit under the Carclo lens of the Astrolux S41? Iā€™m not really sure about it but I like the idea of an S41 Quadruple Red with about 2300-2400 lumens of red light when overdriving the LEDs to 4 amps each. Is this current still okay for frequent usage or do you think thereā€™s a better optimal current/output ratio?

No, the die is smaller than an XM-L but the dome is just as big, doubt it will fit under a Carclo or S41 quad TIR and AFAIK there is also not an XM footsprint triple or quad board available.

Can you remember what the forward voltage was?

I got some from Kaidomain titled ā€œCustom 10W 5050 Redā€ and even though these look like a 100% match, the specs are different in your link. Maybe the AliExpress page has changed over the years to something different?

These ones I have were specā€™d at 3-3.7v forward voltage and yours in the link are 2.2-2.4v.

I asked Kaidomain and they said theirs is a ā€œcustomā€ LED and not Epileds, I think theyā€™re lying.

I agree that the KD version looks like the exact same led as in the OP. The forward voltage can be found in the OP as well, in the table that I made a picture of. (i.e. 2.83 V at 3A). I think that the voltage that KD specifies (3V-3.7V) is not correct and implausible for a red led, almost 1V too high.

Thank you for the reply. I tested it in the Convoy L21B and itā€™s very decent at 3A with 9*7135. I asked Simon to get some and he said he didnā€™t know what these ā€œCustomā€ LEDs were, now I know so hopefully he can get them.

I get 600m and 460 lumens. It provides a good wide beam which is much better than the Osram red

Running on this 9*7135 driver it runs perfectly which has me questioning the forward voltage. Usually with 2.4v it drops battery voltage rather quickly on the Osram red, but this Epileds or custom LED tracked exactly the same as any other 3v using this driver

According to the OP, 9x7135 (3150mA) should measure 2.87V over the led, it will be quite hard on the 7135 chips because unless low drain 18650ā€™s are used, they need to burn off quite some overvoltage. But I have tortured them worse than that and they survived fine.

Using my phone I canā€™t get the correct links to show the graphs

Custom red - Convoy L21B (Custom Red 5050) - Album on Imgur

Osram red - Osram Red 8*7135 - Album on Imgur

Both using a 8x7135 driver for the graphs

Does that look like it has low forward voltage like the osram?

If anything, longer regulation would suggest lower Vf.

Other than them being cheap, I donā€™t think these perform particularly well. XPG3 deep red nearly keeps up at 660nm, which us much less efficient when it comes to radiant power vs lumens.

Also, you can easily measure the Vf with a dmm. Ideally, solder a second long set of wires to the star(they can be very thin since they wont carry any current) and measure the voltage while running. Youā€™ll see the Vf drop at first fairly quickly then stabilize. This is due to the negative temperature coefficient as well ass the 7135s heating up and possibly sagging current

Thank you, just confirmed that it is indeed an Epiled with a Vf of 2.4v on High. Well, that cleared that up, I was hoping to use a regular driver, but it seems Iā€™ll still need to use Convoyā€™s red driver which is limited to 2.5A, maybe I can ask for a 3A version.

I will get some XP-E2 in 620nm and 660nm and see what I think. It looks like they should throw further, but I do quite like the wide beam of the Epiled.

Kaidomain kept denying it was an Epiled until I mentioned I had measured the Vf at 2.4v max, which is when they admitted it, why lie? Very annoying to go as far as listing an incorrect Vf on the website, maybe they didnā€™t knowā€¦

Maybe just use the Osram in a smaller reflector? Something the size of an m21b should be similar to the epileds in the l21b