LED test / review - Daydream custom 110 (4500 K 90 CRI rosy)

LED test / review EN


Daydream custom 110 (4500 K 90 CRI)


2026-05-01


  • Type: multi die, flip chip
  • Bin: —
  • Color group: — (4500 K)
  • CRI: 90
  • Rated voltage: 3 V
  • Max. Forward current: 71000 mA
  • Max. Peak current: — mA
  • Viewing angle: — °
  • Thermal resistance: — K/W
  • Max. Temperature Tj: max. —°C


Toward the end of last year, BLF user billy_s and Chinese driver developer XBY proposed ideas for a high-performance LED with high color rendering and good tint. LEDs of this type are still rare or virtually nonexistent on the market.

Shortly thereafter, a custom order was placed with a manufacturer (which is unknown to me) in Dongguan, China. These LEDs were presented under the name “Daydream”. The manufacturer stated that these LEDs were “not green”, and the first samples were available in early February. The LED tested here comes from this first batch.

The LED has no dome and resembles a ChangDa SFP55. The 25 LED chips are arranged in a 5x5 matrix on the dark gray substrate—in a 25P configuration, resulting in a voltage rating of 3 V. White silicone resin was applied around the chips, and the phosphor was apparently applied afterward as a gel suspension.

According to the manufacturer, 55-mil Sanan chips are used.

The footprint is 110110, with an approximate size of 11x11 mm. Standard MCPCBs with this footprint should fit, but the LED is also supplied mounted on a 32 mm PCB. The thermal pad is stamped with the initials (?) “JL”, which were already present on the NBQ86 I tested earlier.

The phosphor was applied relatively evenly; only on the outer LED chips can minimal differences in the phosphor layer be detected. The particle size of the phosphor is very small, and the size of the light-emitting area is approximately 61 mm².







  • Maximum reached at 60 A, at this point 13764 lm @ 3.32 V
  • Power at maximum 199 W
  • Efficiency at maximum 68.3 lm/W




Data for 25 °C Tsp (at 85 °C the luminance values are around 13 % lower).


The 60 amp power supply was maxed out, so only the maximum current of 60 A is specified.

Performance is very high, as is efficacy. The luminous flux is almost as high as that of the cool-white SFP55, which is likely due primarily to the more efficient Sanan chips (the SFP55 presumably uses chips from AoYang, which have significantly lower efficacy). The actual maximum of the custom 110 is likely between 70 and 80 A.

The SFY55 has 64 chips and a larger footprint, so it pulls far ahead.

The luminance is low. Due to the high color rendering and the large light-emitting area, no more than 110–120 cd/mm² can be expected even at the maximum possible current.

The beam is good. There are no complaints whatsoever, particularly in OP reflectors. There is no tint shift in the beam or corona.



The tint on the sample tested here is excellent, though this always depends on the specific sample. In this custom order in particular, the DUV fluctuated between -0.018 (!) and -0.005. Most samples fell within the range of -0.011 to -0.006.

Several types of phosphors are used: YAG-Ce3+, Ca-a-SiAlON:Eu2+, and likely CASN:Eu2+, although with a presumably lower peak emission wavelength. The cyan gap is relatively pronounced, which is also the reason for the relatively low CRI compared to many other high CRI LEDs (modern high-CRI emitters like B35AM or NTG35).

It is possible that some ß-SiAlON was also used in the 510 nm range to boost the green component and somewhat compensate for the cyan gap without having to resort to more expensive cyan phosphors. However, this remains speculation.


  • Ra: 91.6
  • R9: 90
  • CCT: 4327 K
  • duv: -0.0073

This LED is unique to date. With good color temperature and beam characteristics, very high performance, and compatibility with existing 110110 footprints, it can be easily used in lights, especially at currents around or above 60 A. As long as you can obtain an LED with a DUV between -0.010 and -0.007, it can be absolutely recommended.

Unfortunately, no datasheets or other information are available, and the quantity is limited due to custom orders; furthermore, the variation within a single batch is greater than with other common emitters.


Thank you for reading this test. :slight_smile:

Greetings, Dominik


v1.0.2
21 Thanks

thanks for the Review.
The led looks good and i would buy one to play around with

I really like that color rendering picture! So much info, but the spectrum and r1-r15 are obvious. Is this a custom chart? Or one generated by another program?

From spectrometer PC program

1 Thank

Wow. These look awesome. Thanks for taking the time to test these @koef3
Are they available anywhere?

1 Thank

Interesting…House periphery or driveway lighting might be a think running at around 20A, or 80ish watts…

1 Thank

Nice! Any chance those would be available?

2 Thanks

@billy_s on reddit discord said that the LED is actually 3/6V with one group of 2rows/10 chips and one group of 3 rows/15 chips.

2 Thanks

Don’t think I’m ready to start a dedicated thread so I’ll just borrow Koef’s :slightly_smiling_face:

The 7070 version is here :smiling_face_with_sunglasses:

9 Thanks

:sign_of_the_horns:
This definitely needs it’s own thread!

3 Thanks

I need these!

3 Thanks

Neeeeeeeeeed it!!!
D7f, 3 of these and that gert big xby driver sounds like the combo right?!

4 Thanks

You’re right! :wink:

1 Thank

is it available to buy ?

1 Thank

Yes, XBY have them listed on Goofish.

2 Thanks

Hello. I can’t seem to find a website called goofish? Is it not available in the US? I’d like to order some of these emmiters.

try this. If you really want to purchase something, you may need to install the apk.

2 Thanks

Or

You can paste the link into SuperBuy to order some.

4 Thanks