Yuji VTC-series D50 5730-size midpower led tested

Good news!
Ted Bai from Yujileds just emailed me:

“Hi

I have attached the datasheet of VTC series 135L for your reference .

This is our latest datasheet and not yet been published .

You find all necessary information inside , including the maximum values .”

Yuji VTC 135L-G01

The bad news is, it is indeed low power. But that’s ok, its a beautiful product!

And the most beautiful if you run it exactly @ 450mA, because that’s where it’s calibrated at.

Mid power diodes, I heard someone shout, that’s where we’re going, small size and if you want a lot of light, put a lot of them together on a PCB.

https://store.yujiintl.com/collections/vtc-series

colour shifting from High to low K with 2 dimmers is also a very interesting option with these smashing spectra and boards. Of which the best, the one that looks like a quantum board, ofcourse is sold out lol, but those smaller boards mmmmm

Yuji VTC colormix

The violet spike is somewhat tempered in the new models btw, the gif shows the older spectrum…

:slight_smile:

Finally, I have a very nice super high CRI Yuji floodlight.

The host used is a cheap Krisbow 10 watt floodlight bought from Ace Hardware store. Nothing worth to spare except the case and the driver. This outdoor floodlight has a sleek full cast aluminum body and low profile design. The glass cover is just a simple windows grade glass with sharp edges.

The original MCPCB installed with blob of thermal paste which wan't screwed down properly due to burrs in fixing holes. Total garbage. When I try to de-solder the LEDs off the board, it started to bubble below 220C. I always monitor my hot plate temp.

Found a broken 4 watt bulb, another Krisbow. Took the MCPCB, looked better. The LEDs were 5730 but the solder footprint is slightly different. A little scraping here and there and it fits the Yujis.

It's time to mod the case. Some drillings and M3x0,5 tappings. Swapped the MCPCB from the broken bulbs. Made two slots for screw fixing...

The Yujis installed. Not perfect due to footprint mismatch. I prioritized to align the cathode pads first. The broken path to anode connected with extra solder.

Soaked the finished board in IPA to clean the burnt flux. Tried to brush the brown residue but later changed my mind, too risky.

Applied the trusty MoS2 grease simply because I always forget to buy fancy thermal paste

This time the board is hard pressed against the case/heatsink.

Lapped the back of the glass cover using 320 grit silicone carbide powder to create slightly diffused lens. Looks good!

The driver outputs 9,45 watt to all 9 LEDs @290mA. About 11 watts total power consumed. The case got warm fast but still manageable under 46C (26C ambient).

Comparisons (ceiling bounced)

Thank you for the LED Jos. Now I have a super nice 5000K comparator light

- Clemence

EDIT: In hot sunny day (31C) and still air, the floodlight reached 35C hotter (66C). This worries me for long term continuous use. Need to find lower power supply and possibly better MCPCB. Next will be a fairer comparison to 5000K 9080 E21A

Excellent mod work Clemence!

And nice 4500k and 4000k photos also!

Just got one of these from djozz. I’d say it is quite a nice simulation for daylight.

Testing done at the rated current of 120mA (Vf was 3.48V)

From what I understand about those graphs, that is really impressive.

I have heard that these LED’s can have high levels of UV radiation, what did you see in your testing?

I would not want to use them in a house if they are going to cause UV damage to everything in it over time for example.

The relative UV output at 380nm is an order of magnitude higher than on normal Nichia or Cree emitters. But the X7R for example has as much UV content on low as the Yuji on its rated current.

Man, those are very close to perfect! The have the highest deep-blue cri value (in addition to all the other higher values) which I have seen until now.
Now if only they would start making Power-LEDs. Something like an XP-L with 500 Lumen would be nice. :heart_eyes:

Those are some really impressive CRI numbers. I wish they were targeting the blackbody line instead of D50 though.

Any of those leds left?

Willing to ship to Spain? (I’ll take the extra cost, but please check how much it would be first. It might not be reasonable…)

No sorry, they are no more for sale. I have a few left for my own use, all others were sold.

Ok, thanks.

I personally don’t mind with small amount of (near) UV light emitted by these LEDs. As long as the phosphors still intact I don’t think it would caused ill effect to human. Some incans produced much more UV in the early days.
I found interesting facts about these Yujis. While many blue pump LEDs glow false colour under 365nm UV, Yujis seems to glow it’s true colour. Maybe this is because VTC Yuji is indeed (ultra?) violet pump LED?
This answers why VTC series still limited in the low 60-70 lumens/watt. To become more usable, the (ultra) violet pump LED technology has to evolve to first match what blue pump LED’s achieved. While Yuji uses violet pump, Samsung already started it’s purple pump LED this year.

Note the fluorescence of the transparent epoxy filler in the wood table.


I forgot to test them under blue LED, my bet is the Yuji VTC would glow false colour while the other LEDs would be true

- Clemence

Yes, that is the reason. They have a different type of phosphor (actually three phposphors) for excitation through a UV LED.

But most people don’t run around blasting their X7’s at family and flatmates :wink:

It would be cool to see a 500 or 1000lm (or more) TM-30 Rf 95, CRI-15 95, R9 95+ flashlight, and if that’s too much UV…

I wonder if the Xicato “corrected phosphors” technology also produces increased UV.

High-cri will always be more inefficient because they ouput a full spectrum. The human eye is heavily green biased. The Yuji LEDs emit much more red and deep blue compared to standard LEDs. For the same watt output of light you get less lumens this way. Lumens is basically watts corrected for the sensititvity curve of the human eye.

These 98 CRI, 96 Rf 19mm Xicato Zhaga dropins get 98-130 lm/watt: (3000K) and 88-107 lm/w (4000K). But, the 9mm are less efficient, and they don’t appear to make 5mm versions suitable for a torch.

The link doesn’t work.

says the max for 98 CRI is 100 lm/W and for CRI 83 - 137 lm/W.

Sorry, fixed. Performance characteristics starts on page 13: http://www.xicato.com/sites/default/files/documents/DDS%20XTM%20Artist%20Series%20170221_1.pdf

Those are very large COB LEDs. They have a much, much lower power density (W/mm^2) compared to normal power LEDs we use in flashlights. Power density is what lowers efficency and increases throw.

A Nichia 219B-V1 with R9080 shows what LEDs can do with such high light quality.

These COB LEDs are only suited for floodlights.

Mainly I posted about these because they are very high TM-30 metrics, I know they aren’t designed for portable use.

I realize 19mm = 361mm^2 = much bigger than the 12-49mm^2 for the emitters used in flashlights.

I thought 129 lm/w was a good luminous efficacy for high TM-30 Rf and Rg regardless of emitter area, because lumens are not dependent upon angle/area subsumed. But I could be missing something obvious here.

So the maximum 100% luminous efficacy would be an ideal pure green light emitter, at 683 lm/w.

And the maximum theoretical efficacy of white light emitters is ~ 260-300 lm/w, which Nichia researchers have about achieved, albeit at low lumens: White LEDs with super-high luminous efficacy could satisfy all general lighting needs