LED test / review - Fireflylite FFL351A (3700 K, 95 CRI) - the rosy allrounder

LED Test / review EN


Fireflylite FFL351A (3700 K, 95 CRI)


2024-05-09


For some time now, Fireflylite has been offering special, customized emitters. Some of these are used in their flashlights. The FFL351A is an emitter in 3535 format without a dome, with high color rendering and - according to reports - a slight red tint (“rosy”).

The LED tested here was provided to me by German flashlight forum (TLF) user @Palladin. Many thanks for this at this point!


Technical data


Tj --- °C, If --- mA
  • Type: flip chip, domeless
  • Bin: —
  • Color group: 3700 K
  • CRI: 95
  • Rated voltage: 2.88 V
  • Max. Forward current: 3,000 mA
  • Max. Peak current: — mA
  • Viewing angle: — °
  • Thermal resistance: — K/W
  • Max. Temperature Tj: max. —°C

Note: There is no datasheet available. Values stated above are taken from the item description from Fireflylite homepage.


Design, LES, Performance and Luminance


The FFL351A as such looks quite inconspicuous. The dark gray substrate is covered by two layers, a white and a yellow layer. The luminous surface appears as a darker square area below the phosphor layer. This design is already familiar from other LEDs such as XP-G4 and XP-L2 and has in the past sometimes led to problematic emission properties in secondary optics.

There are no markings on the surface. An ESD element is not recognizable. Although the silicone is quite robust, the LED should always be handled with care due to the lack of a dome. No damage to corners or edges occurred during the test.

The FFL351A is 3.50 x 3.50 mm in size.

The design of this emitter is very similar to the HM-3535 from LatticePower and the XP-G4 HI, although the ESD protection element is clearly visible on the latter.



The footprint reveals no surprises. It is the standard footprint in 3535 format with a dedicated thermal pad in the middle. The clearly recognizable edge in the contact area indicates the cathode (minus).

By using the standard 3535 footprint (also known as XP as known from older Cree emitters), the LED is fully compatible with available accessories, and the symmetrical package means that centering rings (gaskets) produced on a lathe can also be used. This means swapping existing emitters in lights is relatively easy.



The luminous area is 4.3 mm² in size. Some light shimmers through at the side, which makes it much more difficult to determine the illuminated area. This is quite common for modern LEDs in Flip chip design. The LED chip here is probably 2.0 x 2.0 mm in size and thus corresponds to the design of many other current LEDs.



Within official parameters, as far as known:

  • at 3,000 mA (official maximum current): 802 lm @ 3.28 V
  • Power at official maximum: 9.83 W
  • Efficiency at 3,000 mA: 81.6 lm/W
  • Maximum reached at 8.4 A, at this point 1385 lm @ 3.74 V
  • Power at maximum 31.5 W
  • Efficiency at maximum 44.0 lm/W


The general performance of the FFL351A is good, especially when you consider the smaller footprint and high color rendering. A comparison with the FFL505A in particular shows that the larger illuminated area can also make an enormous difference in terms of efficiency. The Vf is very low and is virtually identical to the SFT-40 3000 K, and the overcurrent capability almost fully exploits the 3535 footprint. With this footprint, the maximum achievable current is around 9-10 A, although this also depends heavily on the Vf.

Direct drive of these LEDs - especially when connected in parallel as a triple - is not recommended. Especially when using batteries with low internal resistance and thus good voltage behavior, the LED current can be too high, which massively reduces the service life.

More than 5 A is not recommended for the FFL351A in terms of efficiency and the low luminous flux gain.



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

Despite the lack of a dome, the luminance is not very high. This is primarily due to the low efficiency due to the warm color temperature and high color rendering. The SFT-40 3000 K benefits in particular from better heat dissipation and significantly higher maximum performance. For higher luminance levels, LEDs with a higher color temperature or lower CRI must be used.


Light quality/beam, tint and spectral data



The beam is perfect. There are no disturbing color distortions or artifacts in the beam, despite the areas of the LED chip shining through at the sides. It basically looks like a dedomed 519A, only the corona around the spot might be a little bit bigger.

Slight rings are visible in the beam, which are caused by the reflector used and are less clearly visible in real life. All in all, these LEDs are also suitable for all those who value the best possible beam, especially when OP reflectors or diffuse lenses are used.



The light quality is high. As already reported in advance, the duv is clearly negative, the tint is reddish (also called “rosy”). Ra and R9 are very high and are therefore in good company with the FFL505A. As with other warm white LEDs with very high color rendering, the saturation Rg is also very high, a value above 100 indicates oversaturation, which is mainly due to the high red component.

The CCT is almost 4000 K. Due to the strong red tint, the light appears subjectively warmer, nominally this sample tested here can be classified as 4000 K neutral white.

The light looks very nice, although depending on your own perception, red colors may be overemphasized, which can be particularly noticeable in forests. If you value balanced color representation, you should look for LEDs with duv close to 0 or choose a higher color temperature.


  • Ra: 96
  • R9: 97
  • CCT: 3955 K
  • duv: -0.0064

Conclusion


The FFL351A is an interesting LED. Almost 4000 K, reddish (“rosy”) tint and very high color rendering, very good beam in optics and good overcurrent capability make it a unique LED at the moment, maybe apart from the dedomed Nichia 519A. If modifications to already soldered LEDs are undesirable and a reddish tint is essential, these LEDs are worth a closer look.

In general, these are good all-rounder LEDs that have no real weaknesses and can be used universally for all types of flashlights with 3535 footprint.

Unfortunately, they are only available in China, there is no important information about the manufacturer and - as with many Chinese LEDs - a data sheet is not available.


Pro

  • very high color rendition
  • pleasant tint (“rosy”)
  • very good beam in reflectors
  • good overcurrent possibilities
  • standard footprint for maximum compatibility

Neutral

  • only available in one specific CCT/CRI

Contra

  • only available in China / from chinese sellers
  • no datasheet available
  • no information about manufacturer available

Thank you for reading the test. :slight_smile:

Greetings, Dominik


v1.0.1
11 Thanks

This emitter is very similar to a dedomed 519A 5000K in mechanical/electrical/optical properties, plus CCT, CRI, and tint.

I pulled up your earlier test of the 519A (which happens to be in 5000K!), which indicates that the output of this new Fireflies emitter is between a domed 519A 5000K and its dedomed version, while having a Vf around 0.2V higher on the upper end of drive current. Efficacy should thus be comparable to a dedomed 519A, though with slightly less intensity. I don’t yet see a strong advantage this emitter has over a dedomed 519A.

Thank you Dominic for this test!
If you check temperature with a thermometer is more o less hot than 519 at the same current?

Thank you very much for your FFL519a 3750K test report. Im particularly fond of Rosy emitters and recently bought some FFL519a 3750K, ordered April 13, 2024.

Contrary to posts by several other people, promising Pink Tint, the LEDs I received were not pink at all. I then wrote to Fireflies to ask if they could provide some of the Pink ones. They told me they no longer have Pink FFL351a 3750K

So, people who got them earlier, were lucky. The presently shipping FFL351a are no longer Pink. At least not the two I tested:

There has been a similar shift in the 519a 4500K LEDs that I have bought in the last few months. They have become greener, and even after DeDoming, they barely go below the BBL.

Here is a recent example of a DD 4500K 519a that I put in a Sofirn SC21 Pro:l

I prefer Tint DUV to be lower, and that is why I bought the FFL351a 3750K. Unfortunately, they did not live up to my expectations for negative DUV.

This sort of thing happens a lot. For example, when the sw45k first came out, it Was Very Pink. Each later batch was less Pink.

This also happened with the stock LEDs in the TS10. At first they were very close, to slightly below the BBL. Later batches have become increasingly Greener.

All that to say, i think the Pink Tinted FFL351a 3750k sounded like a great alternative to DD 4500K 519a. However, I was unable to buy any of the Pink tinted FFL351a 3750K. It seems the Pink ones were an earlier production run than those being sold now. Also true about recent 4500K 519a, they have become less Pink.

Fwiw, i tested the FFL351a 3750K in a Sofirn SC21 Pro with a Pebbled Tir. I compared it to an other Sofirn SC21 Pro, also with a Pebbled Tir, using a DeDomed 4500K 519a. fwiw, the FFL351a produced 7% higher output, using the same battery, and same ceiling setting of 120/150.

So, though the FFL351a 3750K had a slight output advantage over DD 4500K 519a, the FFL tint was greener, which for me is not a benefit.

6 Thanks

I did not checked the temps while testing (which is not possible anyway because the LEDs are in the integrating sphere :smiley: ). From the results of my light flux test I think the temps are more or less the same since both Nichia 519A and FFL351A have similar performance characteristics.

Like mentioned by @QReciprocity42 the 519A and FFL351A are very similar, and for everyone who dedomes LEDs, the 519A is also a great choice. But in any case I am happy that the 519A finally got some competitors, after a long time without any.

3 Thanks

This is something many did not understand really well, as they are always talking about “emitter XY has great tint” or “choose these, they are great” - which can be dead wrong depending on the lot the seller has… LED manufacturers give color binnings (or sometimes called “color kits” or “groups” in which the LEDs are sorted in. These tints can vary a lot according to given color binning. This can be totally different from reel to reel, or even from LED to LED.

And my tests are always a snapshot of the current situation, so I am very happy you can explain the differences between lots. My samples are from something beginning of the year or so, so I am not surprised to see a change in tint at all. But apart from that I still think the tint is nice and totally usable, at least from your pics. And it is not so cool from Fireflylite to have still the old spectral data for these emitters in the item description (some would call this mis-leading)

1 Thank

So at today the best efficient LED 3535 is XPL-HI?

Don’t know, I don’t fully tested the XP-L HI (2nd gen, 1st gen is completely outdated)

What I can say for sure, the 2nd gen XP-L HI was completely burned at around 6-8 A, due to burned silicone/phosphor layer - like XHP50.3 HI earlier. So even if these are more efficient, the tint has also to be good (can be difficult with Cree emitters) and also they are not as powerful, the safe maximum current is lower.

Maybe I will make a (incomplete) test of these 2nd gen XP-L HI to clarify this.

the XPL-Hi is not High CRI, comparing it to FFL351a or 519a is not in the same category.

Acc to datasheet, it is available in 90 CRI, but not at the usual resellers I think (only at wholesalers)

thank you, I did not know that.

Thank you Jon but the XPL-HI are with low CRI and with high CRI and i have 3000K 85CRI and they are very very beautiful
The 2850K from Hank are also really beautiful.
The real problem is to find all the beautiful version from seller. Is more easy with 519A or SW27,35,45…
One of KR4 from Hank in the past was equiped with XPL-HI 4500K and trust me, in those version they are very similar to SW45, but it’s impossible to find those LEDs
I try in the hand the same flashlights with XPL-HI and SW45 and the XPL are more efficient and very less hot but yes is very difficult to find their beautiful series LEDs

1 Thank

If you meant efficacy (lumens/watt) I had to guess it might be the 70CRI LH351D: it’s got the largest LES out of any 3535 emitter I’ve seen, and it’s domed. For power efficiency none of them can beat a royal blue LED.

But efficacy is not a great metric as it encourages green tint.

2 Thanks

I meant current/temperature/lumens
Obviously the 6500 are better than 3000

Agree

1 Thank

Very interesting test, thank you Dominik for going through the trouble of testing.

It seems like Fireflylite has been developing a whole lot of emitters, and new color temperatures. Now the FFL351A is also available in 5000k, and very interestingly at 1800k.

I wonder how the 1800k version compares to the Lumileds HL2X in 1800k R8000.

Also, it’s a shame the currently available batch is no longer as rosy as the ones you tested.

1 Thank

Yes I am also interested in 1800K variant but since FFL was not able to ship me some emitters (never arrived despite their tracking says so) I will not order anything from them again… strange that only this order is not arrived, but all others from KD and Ali did…

1 Thank

I’ve had issues with their shipping before, one time with my order of FFL505A the shipping company lost them, and a few other times it has taken unreasonably long to arrive.

I’m guessing you’ve sent Ivy an email? Either way it is very frustrating how bad their shipping provider is.

Tbh I am tired to get answers to orders not arrived… I simply don’t order from them anymore. :smiley: And it seems that I’m not the only one with such an experience. Which Email adress did you use for your complain?

3 Thanks

I’ve also seen the same with the 519A-V1 I purchased from Wurkkos. The 519A-V1 I believe has a bit higher flux than the original 519A which may explain the slightly higher dUV.

I swapped an SC21 Pro LH351D 5000K to a domed 519A-V1 5000K and the tint looked the same to me. Still a R9 gain which is nice. I’m not a fan of rosy so the 519A-V1 domed tint doesn’t bother me.

@koef3 I don’t suppose you have any 519A-V1 to test? Would be interesting to see the measured difference between the original 519A.