Nichia 144A and 229A series: bigger dies, more output, 90CRI included,.......but no thermal slug :-(

start here:

Why the fuss over no thermal pad? If they don’t need it, why do we want it? :confused:

Where the specs stop, we start! :smiley:

Speaking of specs, no link?

added a link. Still have to look at the specs but coming from Nichia they should be good. I’m at work now so I should not spend too much time on BLF :innocent:

It’s a shame, if they made a 5.0*5.0 emitter with high CRI and thermal contact they would sell very well !

Looking over the specs, the 144A is 5.0mm x 5.0mm, and compares to XHP50 in size and performance (but not the top XHP50 bin) with a single chip/die and better color rendering and lower Vf than the XHP50! :partying_face:

Well, it has 90CRI available. And again I ask why we need a thermal pad.

The 90 CRI version of NV4L144AM is available NOW. 3A forward current, thermal resistance only 1.2 ℃/W, 936 lm. Ideal led for a single 18650 “luxury” flashlight.

Erm, not single 18650 unless you have a boost driver. The Vf of the AM is ~6V. The AR is ~12V. Almost like they’re trying to directly compete with XHP-50… :wink:

A thermal pad is needed because you want to thermally connect the led to the flashlight with no non-metal obstructions, and the electrical pads can not be used for a direct metal connection. The output advantage of a direct thermal path only starts at an out-of-spec current (under 3A a XM-L2 shows hardly an advantage) and gets more when current increases.

What would work really well as a non-DTP board would be a non-DTP board like this:

(orange is the copper-poor, blue is the exposed copper for soldering)
If the MCPCB is made with an as thick copper-poor as can be produced, the heat will spread sideways fast and have then enough surface area to cross the dielectric layer efficiently. This board can be used with all the two-pad leds that are out there and boost their performance compared to the common boards.

No-one makes these, perhaps we can ask Hank to have a Noctigon version of this board designed.

Yes indeed. This looks like a drop-in replacement for XHP-50 for the upcoming ZebraLight SC600Fd III Plus.

I think getting those made is a worthy goal. I do understand that a direct thermal path is better technically. But, I haven’t seen proof that Nichia LEDs specifically suffer from NOT having the separate thermal pad. I’m sure it’s impossible to tell, since they don’t make any of their LEDs both ways (with and without DTP pad) for direct comparison. But, what if the emitters that don’t have DTP pads are designed in such a way that heat isn’t the first thing to cause failure? What if removing heat isn’t the bottleneck to performance? I know for Cree LEDs heat is definitely THE issue, but what if for the Nichias it just isn’t?

  1. It’d be nice if it fit on a common Cree footprint so MCPCB’s were easy to come by.
  2. Cree’s probably wouldn’t need a thermal pad at their rated specs either, but because they do we can overdrive them to the levels that make this hobby fun. Likewise I’m sure there is no need for a thermal pad at the Nichia’s rated specs, but it could have been much more fun if there was one. Time/testing will tell.

edit: Late post, Djozz beat me to it.

Direct XML swapping was in my mind…but I didn’t see that the Vf is 6V. It makes the swap more complicated.
Like DavidEF said, it’s more a XHP50 competitor

The 229A is 4.0mm x 4.0mm and ~3V. Not exactly XML size, but could be used for a swap. You’d need to replace (or mod?) the MCPCB anyway, since the Nichia doesn’t have the third pad in the middle for DTP.

There are ways for DTP with custom driver concepts:

a) Reverse the battery so that the body is + and can be electrically connected to LED+.

b) Use a high-side switch, i.e. a P-FET. These typically have a somewhat higher R_DS, but there are very good ones, too. LED- can then electrically connected to body (batt-).

Woa, new NICHIA emitters. Hello High CRI :sunglasses:

Seems like that might lead to the negative side surviving and the positive side turning blue.

So you would use short fat wires from the led into the copper star ?
And heat would go to the star ?
This would work with LEDs that also have a thermal pad. For turbo heat transfer.