LED test / short review - ChangDa SFP55 6500 K 70 CRI - hardly any extra power compared to the SFH55

[koef3 (LED tester)]

LED short test / short review EN

ChangDa SFP55 6500 K 70 CRI

2025-10-29

• Type: multi-die
• Bin: unknown
• Color group: — (6500 K)
• CRI: 70

For further information on the emitter, please refer to [the detailed test](URL zum Test) that has already been carried out.

This emitter was kindly provided to me by BLF user @billy_s. Many thanks for that! I do not have any further information or a data sheet for this LED.

The ChangDa SFP55 is a 25-chip variant of the SFH55, which has already been tested. The “P” in the type designation refers to the 25-chip variant (“H” = 16 chips). Like the SFH55, the chips are mounted on a gray ceramic substrate and covered with a clear silicone layer, under which the yellowish phosphor is located. The chips were probably manufactured by AoYang, which are generally said to have lower efficiency and, above all, overcurrent capability.

The LED measures 11 x 11 mm. This size is usually referred to as “110110,” occasionally (especially on the prints of the MCPCB) as “XHP110.”

• Maximum reached at 60 A, at this point 14393 lm @ 3.27 V
• Power at maximum 196.4 W
• Efficiency at maximum 73.3 lm/W

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

My 60 A power supply was maxed out, which is why the curve ends quite abruptly. The maximum possible operating current should be between 65 and 68 A.

Since the flux binning of the SFP55 (and all other ChangDa LEDs) is unknown, a direct comparison can only be made to a limited extent. The SFH55 is slightly more efficient, despite having fewer LED chips. However, the luminous flux is only about 8-10% higher. However, the SFH55 and especially the SFP55 seem to run massively into their thermal limits; the very small increase in luminous flux indicates that the ceramic substrate is the main limiting factor here. Since a 7070 footprint can already dissipate around 180-200 W, this should apply even more to a larger 110110 footprint.

As expected, the luminance is low; it is further reduced by light emitted to the sides. This LED is therefore more suitable for lamps with a floodlight-like light pattern.

If = 200 mA

With 6500 K and 70 CRI, the SFP55 is a typical cool white LED with low color rendering. Classic YAG:Ce3+ phosphor is used here.

The tint shift with increasing current is hardly noticeable. The light quality is therefore acceptable for flashlights, especially since there is no annoying green tint visible.

• Ra: 72
• R9: -21
• CCT: 7072 K
• duv: 0.0012

The SFP55 is the larger sibling of the SFH55, but it does not manage to truly realize its potential for increased performance. The SFH55 is therefore the more economically sensible option.

Nevertheless, with over 14,000 lumens and a maximum current of well over 60 A, it is an impressive LED that also offers usable light quality within the low color rendering index range, especially with a good beam known from other ChangDa LEDs.

Thank you for reading this test.

Greetings, Dominik

v1.0.2

[Sirstinky]

Thanks for the test! I was waiting for someone to test it. Too bad there’s no datasheet for it. Do yiu think they made a lot of these?

This is the led in the NI03S. Nightwatch says it should do 17,000 Lumens, but I tested it at around 13,100 and 51 amps, so not far off from your results and owing for losses in the reflector, lens.

[Toaster79]

I wanna dIrect drive this thing off an EVE 50PL

[billy_s (K P)]

Even better, Eve 46950

[Wingman]

Your post and that of @Toaster79 above reminded me of an idea I was thinking of a while ago: Ordering a Convoy 3X21D host, but buying the 3X21C driver (which would also fit in the 3X21D), which has a 54A FET mode, so in essence a direct drive mode, although influenced/limited by some ‘bottlenecks’ in the light. Obviously doing all possible mods to as much as resistance within the light as possible. And that “54A FET” isn’t a hard limitation, since I’ve seen a lot of people that have pushed the Amps of Convoy FETs that is stated on the website.

I was thinking of putting either the SFH55 or the SFP55 in it, since it basicly matches the footprint/package size of the SBT90.2 perfectly and so the reflector hole should be big enough for them.

Back then I was thinking about using Ampace JP40’s I believe, since the 50PL (and some others) weren’t available at that time.

@billy_s

Even better, Eve 46950

This was the specific comment just know, that started to make me think: back then, the only option I had where really powerfull 21700’s. But in the meantime I’ve bought several 46950 rewrapped by Liitokala, but after a bit of investigative work and taking measurements of certain things which where all perfectly within the specs that EVE lists on their official datasheet (all of this I reported here on the forum), I’m 99% certain that the come out of the EVE factory.

Because I know had a bunch of 46950 cells and no real purpose for them, I bought a 46950 battery tube from Pioneman (before Simon started selling them too, for halve the price… ) because I expected to buy a 3x21C or 3x21D anyway at some point, but so far it hasn’t happened.

When you mentioned using the 46950 in combination with the SFP55, my old “plan” popped up again and over the past hour I started thinking again about it (and looking up some current prices for all the parts), but this time with the difference of using a 46950 instead of 3 50PL (of which I have 3 new and closely matched, in storage), to finally maybe doing this project and FINALLY using that battery tube and 46950 in this light!

[Toaster79]

And the “driver” for it

[billy_s (K P)]

@Wingman

When you mentioned using the 46950 in combination with the SFP55, my old “plan” popped up again and over the past hour I started thinking again about it

But, to mix things up, DDH now offers a very good looking 3x21700 tube.

[billy_s (K P)]

Can we call that a driver tho

[tinky]

Hahahaha, hilarious

[Toaster79]

All you need is a momentary switch to turn it on. Yeah, it’s a driver also has a tiny led on it so you can see it’s on . Definitely a driver

[Wingman]

@koef3 I don’t know if you’ve measured/calculated the illuminated area of the SFP55?

If not: do you know if the individual LED chips of the SFH55 and SFP55 are the same (size)?

That way I could (roughly) calculate/estimate the illuminated area of the SFP55. If indeed the same, I roughly come to 49,5-49,9 sq mm vs the 31,9 sq mm you measured for the SFH55.

[koef3 (LED tester)]

@Wingman

If not: do you know if the individual LED chips of the SFH55 and SFP55 are the same (size)?

I am currently not at home and have no access to LEDs and equipment. Please measure the LES by yourself since the dimensions are mentioned in the test.

[Wingman]

I’m sorry, I missed the part about the 11x11mm package size, sorry for asking/bothering you!

[INeedMoreLumens]

@Wingman

do you know if the individual LED chips of the SFH55 and SFP55 are the same

Yes, the 55 means the individual chips are 55mil.

3v:

6v/12v:

[koef3 (LED tester)]

@Wingman

I’m sorry, I missed the part about the 11x11mm package size, sorry for asking/bothering you!

no problem next week I have more time if you have some questions.

edit: sorry, misunderstood your question… yes, the individual LED chips are same size, as @INeedMoreLumens correctly pointed out. Only the amount of chips varies between the two emitters…

[Wingman]

@INeedMoreLumens

Yes, the 55 means the individual chips are 55mil.

Ouch, I totally forgot about that!

[cannga]

Regarding “ChangDa SFP55 6500 K 70 CRI - hardly any extra power compared to the SFH55”

Is the SFH55 and SFP55 you tested the same as the SFH55 & SFP55 listed by Nightwatch on the sheet below? Per this sheet SFP55 has a lot higher output than SFH55, no? Or is SFP55.2L that much different from SFP55?

I have lights with both LEDs: Mateminco PD90S with SFH55 and Nightwatch N103S with SFP55.2L. At least with these lights my measurements are consistent with the Nightwatch sheet: SFP55.2L has much higher output.

[koef3 (LED tester)]

@cannga

Is the SFH55 and SFP55 you tested the same as the SFH55 & SFP55 listed by Nightwatch on the sheet below? Per this sheet SFP55 has a lot higher output than SFH55, no? Or is SFP55.2L that much different from SFP55?

Are the values from Nightwatch?

From my experience with other LEDs with measuremens by Nightwatch (or whoever these measures) there were always differences, my measurements were consistently 20-30 % lower. I even checked my sphere and ran a full calibration on it after discovering this. Since all my measurements where consistently plausible (for both flux bin and claimed values) except for these LEDs I don’t think that my sphere or equipment is measuring that wrong.

[cannga]

I don’t doubt your measurements for a second. BTW this puzzle doesn’t matter much to me except it is always interesting to play detective with “Chinese LED” mysteries .

1. Nightwatch numbers: Agreed that numbers from Nightwatch are inflated, but they would then be inflated for both LEDs, therefore wouldn’t explain the discrepency. Nightwatch numbers show SFP55 has nearly twice as much output (17000/9100=1.87)

2. Actual measurements in flashlights:
   SFH55 in Mateminco PD90S, 1lumen review here: ~6600 lm Turbo turn on. My measurement is identical FWIW.
   SFP55 in Nightwatch N103S, 1lumen review here by @Sirstinky: Nick and I got identical 13000 lm Turbo turn-on.

Essentially both Nightwatch and actual measurements in flashlights show SFH55 in the 6000-7000 range and half that of SFP55. I still keep thinking I am erroneous with this (amateur) analysis somewhere? Or did they “slip” you a new/higher power version of SFH55?

[elemental]

I would strongly suspect the Mateminco PD90S’s driver or overall current path has some hidden resistance that is ultimately limiting its performance. The Nightwatch driver is much better in that regard. It would be more fair if they were compared within the same driver/host so that we could rule that out.