Alibaba GT-FC40 tested (a 5000K 95CRI 12V domeless XHP70 replacement, supplied by rngwn)

BLF member rngwn sent me two of these interesting 7070 leds for testing. He found them on alibaba and asked for a few very high CRI samples. Depending on the performance he will order a batch of these and sell them to BLF members.

I did not get the manufacturer's name, or a datasheet, just the typenr GT-FC40, and that they are 12V, 5000K, and very high CRI.


First how does the led look? It is clearly a 16 die led (4 strings parallel of 4 dies in series, I assume), but very tightly packed so it may produce a good beam in smooth optics, just like the XHP35.Here is a picture at 1 mA, to show the emitting area.

Not all dies show the exact same emitting colour. This is not because of the very low current, at 1A the same colour variation can be seen. Also, see the top pictures, there are some specs of dirt in the led, they are not on top but mixed in the phosfor. I hope that production batches do not have that, but at least the performance does not seem to suffer, as seen below. All in all production control seems not perfect.


I measured the tint of both samples at 1A, straight above the led, and at 60 degrees. Then I placed a small smooth reflector on top of led nr. 2 and measured 3 positions in the beam: middle of the hotspot, the corona, and the spill. Here is the tint data of the hotspot from the smooth reflector:

The beam looks really good by the way. No obvious ugly tint shifts, just a mild peachy corona. Here's the beam of led nr2 at 1A through the smooth reflector, projected on the ceiling, underexposed to show the profile best. That looks very pretty I think!

7 measurements is too many to show all the graphs, so here is a summary of the tint data in table-form.

CCT CRI R9 duv
led 1 bare, straight above LES 5340 95.5 93.8 -0.0035
led 1 bare, at 60 degrees 5080 94.6 85.2 -0.0007
led 2 bare, straight above LES 5410 95.6 97.8 -0.0049
led 2 bare, at 60 degrees 5150 95.1 90.6 -0.0023
led 2 in smooth reflector, center hotpot 6220 94.9 92.4 -0.0023
led 2 in smooth reflector, corona 4780 95.6 90.6 -0.0018
led 2 in smooth reflector, spill 5248 95.6 93.7 -0.0035

Now that looks like the manufacturer delivered what was asked for, a solid 95 CRI in both leds at various circumstances, and everything close to but below the BBL. No complaining here.

Output performance

This is a 12V led, but the best comparison is with a XHP70, with dome ibtact, it more or less shares the same die-area.

So in the graph I added my only, pretty old, test data of a XHP70 led, a N2 bin led while a modern XHP70.2 led is at bin P2 which is 3 bins more output. Nevertheles it will give an impression where this led fits in. Mind the voltages btw, the tested XHP70 is a 6V led so not 1 to 1 comparable that to the 12V FC40.

It looks like the FC40 leds perform similar to the old XHP70 at the same current, but at double the voltage, so almost twice the wattage / (a bit more than) half the efficiency. The lower efficiency is normal from a 95 CRI led, but I am pleasantly surprised by the power handling of this led (at 7A it ran at 92W for a few minutes), and the top output of over 4500 lumen (although I would drive this led at maximum 4.5 A / 3500 lumen)


I like this led, it looks good, gives a good beam, has great CRI and has ok efficiency for this great tint..But it is 12V, so not easy to drive. A 4A 12V driver is not common, big and also expensive, so it is a bit exclusive to use this led in a flashlight build. I hope that rngwn manages to have the manufacturer a 6V version built, or even 3V !

The tint seems amazing. I hope the warmer temps will have similar DUV values. Would love to buy several of these on 16-18mm mcpcb to go inside the Thrunite T1 that comes with a very nice 12V driver.

If this can be made in 3V and 6V would make this much more practical. Maybe rngwn can send this test report to the manufacturer.

Thanks djozz for testing these!

Very interesting LED.

It is interesting that all the tints look good except the hotspot in the reflector being 6200k.

On paper it has a lot of tint shift in a reflector, is it noticeable in real life?

Besides that and the low efficiency it looks really nice.

I agree, a 3V version would be amazing.

Thanks for testing, interesting for sure!

Can you desolder one and give us a look at the footprint of the led?

Thanks to rngwn for supplying djozz with these leds to test and to djozz for doing the testing. I still use your charts djozz for comparing leds. :+1:

Thanks all of you for the appreciation! And nice to see you are still around Steve :slight_smile:

Thank you for the tests djozz.
Nice CRI and Duv but they should improve the manufacturing process.

Thanks djozz, looks like a winner. :slight_smile:
This led is made by Getian. I wonder how it throws…it’s probably flip-chip so may be worse than wecould hope for.
Looks like a great use for the loneoceans 100w boost driver. :slight_smile:
I also wonder about the performance of lowers CRI variants….

The 70 CRI 8000-9000k is quoted nominally (at 3A) at 3000-4000lm. That should translate to 3500-4500lm at 5000-6000k CCT.

The one Djozz tested is quoted at 3000-3200lm @3A.

With Simon’s 12V driver it should yield more than 2000 high CRI lumens on turbo. I’d be satisfied with that for now until something better comes along. Thank you for the test, djozz!

Good work Jos… :wink:

Steve’s waiting around for you to finish the flashlight so he can provide his address… :stuck_out_tongue:


If this led was in 3V I would have many ideas for it and probably the first flashlight already built.

Edit: it is at least not a design problem, looking at how the led is build up for 12V, a 3V version would simply look like this:

I unsoldered the led to have a look at the footprint. It looks like it uses a board made for XHP70 in 12V configuration. I checked continuity of the thermal pad and the core and there is, so the board is DTP: the entire thermal pad seems one with the core.

The footprint looks like its for 12v only. But at least you can swap the MCPCB size.

So, what happened to this LED? No news in over a year…

Still for sale, still a nice led.
There are some reports of the use of this led in rngwn’s sales thread: WTS/GB: [4000k ARRIVED, Test Report's In] GT-FC40 (a.k.a. XHP70's flattop, 12v 40w sister) 95 CRI, Available in 4000k, 5000k

Simon also got samples of this LED from Getian, but no further details. I think he said he used to work for this company at some point in time too.

I took some beamshots