NMD: I put one of the newly released XHP35.2 Pro9 emitters in my KR1. All my beamshots sucked, and I don’t have any fancy measuring equipment, so please enjoy these graphs and words.

**Background**: Cree has put out another variant of the XHP35.2 they call the “Pro9” ([press release](https://www.cree-led.com/news/pro9-leds-offer-breakthroughs-in-efficacy-and-reliability/), and [datasheet](https://downloads.cree-led.com/files/ds/x/XLamp-XHP35.2-Pro9.pdf)). The selling point is that these Pro9 LEDs are only in 90 CRI and 95 CRI and supposedly have “15% higher efficiency” compared to the non-Pro9 90/95 CRI variants. This intrigued me. However, there is one caveat: the Pro9 is rated for 1/2 of the maximum current (700mA vs. 1500mAh at 12V). I’m dumb so I swapped the GT FC-40 in my KR1 for these Pro9 things in 4000k with 90 CRI. **SPOILER**: despite exceeding the maximum rated current, it didn’t explode.

All graphs were made with the stock thermal limit of 45 degrees and are on TURBO.

​

[KR1 with XHP35.2 Pro9 ceiling bounce](https://preview.redd.it/9bop1kjp17rc1.png?width=1400&format=png&auto=webp&s=d1b5f9e0add7597f6ec7d13b973b3838bb3e737c)

​

[KR1 with its original FC-40 \(ignore the blip at the end\)](https://preview.redd.it/l7ttm6i127rc1.png?width=1400&format=png&auto=webp&s=05a9c51c85349338778a8f457fd83688f421ef54)

​

Since I lack an Opple and integrating sphere, ceiling bounce graph are all the science I can do. However, I think the runtime patterns are kinda interesting.

​

It seems like the Pro9 does not like to b(e overdriven. In the first 30 seconds of runtime it drops 38% from it’s initial value at turn on. Ooof. The FC-40 only loses about 16% from its time zero output level. This not great. Even the Convoy drivers (which apparently by design ramp down immediately upon startup by design) don’t dump output this quickly. I am not a scientist, but my uneducated guess is that all the extra red they added to phosphor causes the junction temperature to get hot real quick, which results in thermal resistance killing the output. Hopefully someone who knows something can explain it.

**Subjective Impressions**:

* *Brightness*: The brightness on this emitter is “OK”. It’s got more punch than the B35AM in my S3. My unscientific testing using the Ceiling Bounce app in combination with my Integrating Closet Shelf (it’s a white shelf in my closet), the Pro9 generates lux values well above what the B35AM in my S3 does, somewhat above the W2 in my L17, and little below what the 6500k SFT-40 with a Convoy 3V 8A driver generates (for a couple seconds). At 30 seconds it closer to whatever the L17 does. My guesstimate is in the approximate range of 1500 lumens.
* *Tint/Colour*: The tint bins for this emitter make positive DUV quite likely. My sample is slightly green. It isn’t too bad. The CCT/tint of the emitter changes A LOT as you get higher up on the ramp. At ceiling it looks much cooler and more neutral. On turbo it becomes the purest white light I’ve seen. The colour rending is OK. On the R*o* (that’s “R orange cat”, because my cat is the best/most convenient way assessing an emitter’s ability to render red) I would rate it “pretty decent”. It’s better than the LH351Ds that I have kickin’ around, and not horribly deficient compared to the 519as I have, but it is definitely inferior to my samples of 219b, and the GT FC-40 that was in this light originally.

​

**TL;DR**: Cree has a high-CRI only variant of the XHP35.2. It is rated for much less current than the regular version. It doesn’t blow up when you put it in a Hank Light. Tint is not great. Turbo-ing makes the emitter unhappy. It might be OK if I used it for its intended purpose (low to moderate output), but that isn’t fun.

​

by g_buster