Cree XP-P new throw king?

Hey folks this is my initial testing of the Cree XP-P, i don’t have a lumen tube so i am doing what i can to measure output in different ways.

I bought two Cree XP-P from Cutter both are on 20mm T-Pads both are U5 bins, the XP-P i am testing is the 5700k the other XP-P i bought is NW 5000k.

Cree rate the XP-P at 700 lumens at 3amps, in my testing i got 750 lumens @ 4.5amps but that is inside a light. Throw is better than the best XPG2 dedomed but output being only 750 lumens is a tad down, if compared to a XPE its a whole new level.

I tested the Cree XP-P in a C8 using a Convoy 20mm 5 amp driver and a high drain cell, after checking if it would burn out with a low drain cell first lol. The XP-P tested at 4.5 amps max current for a whopping 750 lumens.

First look Cree XP-P left and Osram KW CSLNM1.TG

!!

!!

!!

The XP-P looks happy at about 4.3-4.5 amps
!!

I measure the lux at 5 meters indoors and i took the average of the 3 readings which was 8130 lux @ 5m 203250cd

The three reading are as follows

  • 7750 lux a @5m = 193750
  • 8540 lux @ 5m = 213500
  • 8110 lux @ 5m = 202750
  • Average 203250cd

Fence is 20m away
!!

!!

!!

!!

!!

Tress are 150m away

!!

!!

!!

For comparison i have a C8 with the SFT-40 and you can visually see the throw difference i ttesed this C8 at 169kcd and 2400 lumens

!!

!!

Nice review. What does a c8 do (LUX wise) with a boost HL and boost HX

And also what is the LES (light emitting surface) of this new cree XP-P?

LES is 1.05mm*1.05mm, so 1.1mm².

On a lm/mm² basis, it should underperform 1mm Osrams by a measurable margin and be a near match to PM1s. But real world results are what really matter in the end.

I’m curious what numbers you get with NM1 in the same host as this XPP sample.

I have a C8+ I built with a 5A Convoy driver running a 1mm white flat which is putting out 230Kcd. I was hoping these new emitters were going to be up there with that, but with better tint!

If i remember correctly the Osram KW CSLPM1.TG does around 168kcd in a C8 i think.

The smaller Osram KW CSLNM1.TG was doing just under 200kcd in a C8 i don’t have it any more as focus was rubbish. When using the 3030 centering ring both the CSLNM1.TG and XP-P perform worse than using a 3535 centering ring.

Isn’t XP-P 3535 emitter?

If so, no wonder you have focus problem

Not sure cause the 3030 where made for the Osram CSLNM1.TG and they fit the XP-P fine just don’t work great, its most likely the height that the reflector is sitting at these LEDs are really low. The 3535 centering ring is to big for the XP-P.

The osram emitters are thinner than most, so they require a thinner gasket. Early 3030 gaskets like ones I sourced from mountain where waaayy too thick. With the HHL532, which is an even lower profile chip, usually tunes in at 0.40mm IIRC in convoy reflectors. So try 0.60mm and sand slowly from there

Yea i use to do this all the time when the XPL-HI came out but then i stopped because some of the centering rings where not flat after sanding lol. I own a sanding now.

Looks like Cree is introducing these in all CCT and CRI options based on their data sheets which is amazing. I have yet to find a place to buy warm white 90 cri variants.

Very interested in the high CRI versions, especially the 5700K.

Inspired by all the testing above (and beam shots, love a good beam shot!) i just did a very quick, very crude test to measure an XP-P against a CULNM1 (4040 1mm2) in a Cometa. My interest is for lower current lights, especially single AA, and especially those with custom 3535 MCPCBs.

The Cometa was already setup with the CULNM1 being powered by an L4P LD-4 driver, i set the driver to 4A and took measurements for modes 3, 4 and 5, then swapped in the XP-P and repeated the measurements.

Tailcap current measured the same for both LEDs with a fully charged battery so i tested with one at a 3.88V to get an idea of the real world effect of the XP-Ps higher Vf.
With one exception (XP-P on max) the currents provided were the same with both batteries, and i have assumed the lux would stay the same too.

Lux readings were taking at a random but importantly consistent distance (around 5m) so are for comparison between the LEDs only.

With 3.88V battery voltage:

Mode 3:
CULNM1: 0.375A - 1840cd
XP-P: 0.375A - 2060cd (12% increase)

Mode 4:
CULNM1: 1.177A - 4700cd
XP-P: 1.177A - 4910cd (4.5% increase)

Mode 5:
CULNM1: 3.97A - 11100cd
XP-P: 3.52A - 7760cd (31% decrease)

Battery voltage at end of test: 3.87V.

Mode 5 repeated with 4.15V battery voltage:
XP-P: 3.97A - 8000cd by the time i measured but dropping rapidly, down to about 6500cd in 30 seconds.

In mode 5 i think the XP-P suffered from serious heat sag, i didn’t use thermal paste and it’s possible the retaining ring wasn’t putting enough pressure on the MCPCB for proper heat transfer. (The current remained the same.)
So it’s possible the XP-P has more to give at 4A but given JaredM’s measurements and the effect of the higher Vf i would still expect it to underperform the CULNM1 at 4A.

Up to 1.177A though, with a 3.88V battery in this test it outperformed the CULNM1 which is enough to make me look at further testing at lower currents, maybe up to 2A.

And we have emitter performance variation variable where it is quite possible that you got lower performing Osram and higher performing Cree on a test.
Osram can take some serious abusing while Cree not. I personally tested Osram in FET driver setups and I can tell that it is almost indestructible. Although one must be more carefull when doing FET driver diy setup and not cross 5A current draw if possible. FET practically means direct drive so performance depends on used battery and it will constantly dropping from 5A current.
FET driver is still The best and the coolest(coldest running) setup in my book because Osram emitter does not loose much performance with battery voltage sag and whole setup runs much cooler than linear or buck/boost type of driver with constant current draw. Current draw(battery sag) in such setup is proportional to temperature loss. 4.2V, 5A, 55°C > 3.3V, 2.5A, 30°C

Since I last checked it seems there are now more of the options available for CCT/CRI (I’m lookong at digikey). Has anyone tried the 3000k, 90 cri yet? Looks to be in stock.

dumb question - but doesn’t ‘‘throwiness’’ depend on optics, more than the LED?

wle

It depends on the optics/reflector but also on the size of the light emitting surface, the smaller - the throwier (see Osram W1).

It depends on both, a larger-LES emitter in the same optics will have a smaller hotspot than a large one, and also will probably have lower intensity, because the focused light is spread wider. The ideal thrower LED would be as close to a point source as possible, because the light intensity (lumens/mm²) would be much higher, it is why HID and LEP throwers are the highest intensity lights available.

Compare for example a Cree XPL HI (3.7mm²) vs an Osram W1 (1mm²): the XP-L can output more than twice the lumens but it only has a die intensity of 430lm/mm² compared to the incredible 890lm/mm² of a high-bin W1.

The higher instensity means it’s possible to achieve a lot more throw using the same size optic/reflector, or the same throw with a much smaller diameter.

I got some 3000k ,90 cri from digikey. They are very nice! I don’t have a way of measuring, but to my green-hating eyes these look to be very near the bbl if not slightly negative duv.

What may keep these from being the perfect warm, high cri throwy led is the low max current. When I find some time I plan to do a triple and a dual-channel D4V2. Multiple emitters will help with the current/output limitation.

I ended up putting these 3000k 90 cri in a triple wowtac a6 and half of a d4v2 dual channel. The beam profiles in both lights leave much to be desired but I want to try adjusting the height of the optics to see if that helps. For the time being they both have dc fix which gives a nice beam but does cut down the intensity.

Tint on these is fantastic. I’m going to try the 80 cri as well. I’ll also try out some 4000k but I just have a bad feeling they’ll be greenish. Has anyone tried those?