Lumileds Luxeon V2, a 5000K 70CRI led (from Arrow) tested.

djozz · 2019-01-04T09:47:21.000Z

Cereal_killer:

Wow, I was looking forward to this test but I must say the results make me sad, I can’t believe just how much sanding down the diameter of a carlco quad optic a few mm really costs… I may as well swap to a triple, ~~probably~~ for sure I’ll get more lumens otf at lower current.

I’m seeing >40% optic loss right now!!!

My djozz-lumen most likely is between 7% and 10% high (depending on actual spectrum) compared to real lumens (I accept maukka’s calibration as closest to real), but I stick to my djozz-calibration for comparison reasons to older tests. That said, the djozz-calibration is way lower than what most flashlight manufacturers (including the high end ones) seem to use.

USA · 2019-01-06T14:28:00.000Z

Thanks for the build report and the measurements. I also believe this OTF loss is on the higher end, but I don’t have anything to compare the result to.

I am planning on using the Carclo quad optic for the V2s that might come in tomorrow.

djozz · 2019-01-06T15:52:46.000Z

38% loss from a XP-type led in a Carclo optic sounds like something is wrong. I do not see that the Luxeon V2 is optically so different to account for that. A XP-G2 shows about 18-20% loss in a Carclo 10570.

EasyB · 2019-01-06T16:27:18.000Z

These style LEDs that have the phosphor to the side of the die also emit significant light from the side-phosphor. The TIRs might not collect that light as well so that could explain the additional loss.

contactcr · 2019-01-06T19:06:35.000Z

I do not buy these optic losses. There is likely some other variable at play.

I do not trust my sphere cause the napkin math doesn’t make sense but here’s what I get when I take apart my Emisar D4 w/ L1V2’s:

2,650 lux w/ optics + lens (143lm)
2,780 lux w/ optics no lens (150lm)
2,870 lux w/ no optics no lens (155lm)

I used the linear setting it defaults to when you reset the battery. Granted this is the 10621 optic both show like ~90% efficiency according to their respective datasheets.

tvizk · 2019-01-06T23:37:42.000Z

contactcr:

I do not buy these optic losses. There is likely some other variable at play.

I do not trust my sphere cause the napkin math doesn’t make sense but here’s what I get when I take apart my Emisar D4 w/ L1V2’s:

2,650 lux w/ optics + lens (143lm)
2,780 lux w/ optics no lens (150lm)
2,870 lux w/ no optics no lens (155lm)

I used the linear setting it defaults to when you reset the battery. Granted this is the 10621 optic both show like ~90% efficiency according to their respective datasheets.

Yea honestly I was surprised by it too. I’ll try it without the lens later and see what I get. The lens isn’t AR coated.

contactcr · 2019-01-07T00:09:47.000Z

What setup are you using to measure? I think your lumen reading with no optic maybe be getting direct light to sensor when measuring “OTF”. The beam angle is much wider after all so it can spill into something very near it.

tvizk · 2019-01-07T00:12:33.000Z

contactcr:

What setup are you using to measure? I think your lumen reading with no optic maybe be getting direct light to sensor when measuring “OTF”. The beam angle is much wider after all so it can spill into something very near it.

Just a box with paper lined in it and Zak’s ceilingbounce app. That’s a good point. Unfortunately I don’t have a DMM at the moment to measure the current draw, but it’s got 20awg driver leads and bypass with a solid copper heatsink and with a fresh VTC6, at 5200 lumens seems to pretty much drive the emitters to the max.

contactcr · 2019-01-07T00:22:56.000Z

That number is too high for a triple and I’ve had similar issues with my shoe box setup. Try adding a small white divider between your phone sensor and the light:

Even this is probably not enough but I suspect you will see lower numbers without optic and similar numbers with optic. If that’s the case you had too much light getting to the sensor without bouncing around.

tvizk · 2019-01-07T00:56:42.000Z

contactcr:

That number is too high for a triple and I’ve had similar issues with my shoe box setup. Try adding a small white divider between your phone sensor and the light:

Even this is probably not enough but I suspect you will see lower numbers without optic and similar numbers with optic. If that’s the case you had too much light getting to the sensor without bouncing around.

Alright so, I just remeasured and did it with 2 phones as well as a divider between the phone and light. With one, I got 2700 with optics and lens, 3591 bare. On another, I got 2900 and 3790. So that’s a 23% and 24% loss. I think that seems right? This was done with a VTC6 at about 70% charge.

contactcr · 2019-01-07T01:00:55.000Z

That is much closer to reality I think.

Lets say you are pulling 12Amps at 0s = 4200 djozzlumens * .80 estimated losses = 3360 djozzlumens at 0s, remove 7-10% for djozz bragging factor = 3024

Is this even english?

tvizk · 2019-01-07T01:06:27.000Z

contactcr:

That is much closer to reality I think.

Lets say you are pulling 12Amps at 0s = 4200 djozzlumens * .80 estimated losses = 3360 djozzlumens at 0s, remove 7-10% for djozz bragging factor = 3024

Is this even english?

I need to get a DMM. Anyway, I’m still happy with the Luxeon V2, although I don’t actually know what to do with it. I just saw Luxeon and 3535 and ordered because I haven’t actually seen one that is bad. The tint on this is absolutely gorgeous, very similar to the X-PL HI 3A, but less rosy.

contactcr · 2019-01-07T01:07:21.000Z

If you decide the divider is more accurate I would re-calculate back to some of your old measurements for the “with optics” numbers so you have a more accurate “floody mule” number but similar numbers for everything else. If that makes sense. In reality it’s probably something inbetween but you may not want to try and redo a whole new baseline just for this.

tvizk · 2019-01-07T01:09:22.000Z

contactcr:

If you decide the divider is more accurate I would re-calculate back to some of your old measurements for the “with optics” numbers so you have a more accurate “floody mule” number but similar numbers for everything else. If that makes sense. In reality it’s probably something inbetween but you may not want to try and redo a whole new baseline just for this.

Yea I’ll probably stick to this. I don’t have any light that uses otpics for now, sold all of them since I find them too floody. The lights with reflectors don’t have much difference in measurements with the divider so all is good.

contactcr · 2019-01-07T01:12:31.000Z

tvizk:

I need to get a DMM

If you do get the UT210E. I think it can be had with a coupon for ~$30 from Banggood. (contact your preferred affiliate spammer for details )

The Luxeon V2 4K is great too. I have it in a BLF A6 and I have a mixed D4 with 2 of each.

tvizk · 2019-01-07T01:14:47.000Z

contactcr:

tvizk:

I need to get a DMM

If you do get the UT210E. I think it can be had with a coupon for ~$30 from Banggood. (contact your preferred affiliate spammer for details )

The Luxeon V2 4K is great too. I have it in a BLF A6 and I have a mixed D4 with 2 of each.

Haha, that’s actually exactly what I have in my cart right now and am going to order it!

Ryley · 2019-01-07T05:37:33.000Z

Here’s mine, installed in S2+ with frosted 15 degree TIR and compare with Samsung LH351D 5000K 90CRI. Photo taken by cell phone using Fluorescent WB.

Samsung (left) - L1V2 (right)

kikkoman · 2019-01-07T13:51:21.000Z

@djozz,

was there any difference in the Vf measurement method (since the Oslon mentions ‘V over LED wires’ and the others don’t)?

djozz · 2019-01-07T14:05:46.000Z

kikkoman:

@djozz,

was there any difference in the Vf measurement method (since the Oslon mentions ‘V over LED wires’ and the others don’t)?

No, the Vf is always measured the same way: ~5cm long 20AWG teflon wires were soldered to the ledboard, the the wires for voltage measurment are clipped at the end of those wires. Ideally you want to solder separate voltage wires right next to the led but I never do that. The voltage drop over those wires is small but not negligable: at 5A it is about 0.03V, this should ideally be substracted from the voltage value.

tvizk · 2019-01-08T14:20:39.000Z

Got another light to try out the Luxeon V2 again, this time in 3000k flavour.

And as usual, Lumileds did not disappoint. Beamshot was taken with WB4600, S1/15 and ISO100 which is pretty close to what I’m seeing.