Testing a Cree XHP50.2 J4 3A led

I have a dumb question about the LED. I want to reflow one and can’t figure out where is + and - :person_facepalming:

Dang dunno offhand, but check the CREE datasheet and usually there's a mark on the underside to indicate anode or cathode side - I always forget which is + and - and always looked it up. Finally wrote it down, but my notes are @home . I sanity check all the new LED's now - used to be easier.


This should solve that, but it is not how it looks, I soldered the led the wrong way around on this board :person_facepalming:

So I say the round dot is + and the square dot is -

If you look at picture in the OP, the side with the square is the cathode (-) side and the side with the circle is the anode (+) side.

EDIT : beaten by djozz :person_facepalming:

Put the two dots on the bottom, plus is on the right. :wink:

This is on the XHP-50.2 ONLY! Most modern Cree emitters have the plus on the side of the bond wires. The XHP series makes it more difficult to tell but of course you can always check the data sheet, way at the bottom end.

I have a xhp50.2 on the way. I will do a direct comparison of beam size and shape with my sliced xhp70 in my L6 with SMO reflector.

I’m curious which will be the better thrower.

I have a question. I have a TN31 modded with a K40M driver (resistor mod) and a dedomed MT-G2 Copper added to the LED side and driver cavity + spring baypass. Would a XHP50.2 be an upgrade (throw or total output)?

XHP50.2 is on the way, can’t wait to mod it to my nitecore EC4

The more amps those 2 see the more the 50.2 pulls away. At 5 amps for instance, the MG-T2 does about 2800 lumen and the 50.2 does 3500. So I’m sure you would gain brightness.

Since the 50.2 is a smaller die, you would probably gain throw as well.

In the areas of tint, my 5700k 50.2 has a pretty noticable yellow corona (looks square shaped as well). I just installed it though, so I’ll need to try it out at night and see how it looks. In my Convoy L6 with lowered smooth reflector I do see a slightly dark center. No where near as bad as my sliced dome xhp70, though. Reflector height plays a large factor here.

So yeah, I would think it is a good upgrade to the MG-T2.

The jury is still out whether it’s better or not than my sliced dome xhp70. I’ll post up hot spot size comparisons tonight.

Here is my light. A Convoy L6 with an SMO reflector that has the bottom removed. This drops the reflector down just under a millimeter (roughly). I use very thin copper shims to raise the emitter to fine tune the focus. Right now I’m using 2 shims. I reduced the camera ISO to better see the beam shape.

Here are the shots all from the same distance.

Now here is the same with me raising the reflector slightly. This would be closer to a stock SMO.

You can see that the first pic has a less defined edge on the hot spot, a smaller corona and a slight donut hole shape.

The second pic has a more defined edge, larger corona and less donut shape. This looks better in real life, but does not deliver the highest lux. If your chasing lux then the first picture will give you the highest readings.

Here are the same basic pictures as above taken with a more realistic ISO setting.

In real life the donut hole in the first pic is quite subdued, but you kind of notice it a little as you move the light around. The second pic it’s less noticable and a nicer looking beam in general.

I will compare this xhp50.2 to my sliced dome xhp70 in just a little while. One interesting thing is that the corona on the xhp50.2 is noticably square shaped at close range. It’s kinda odd. Lol

Here is a pic of the same light with the CW xhp70 in it.

You can see the left pic has a smaller hotspot. I’m not sure if that’s due to the reflector going closer or the dome being sliced off. Maybe the dome. It looks pretty much the same size as the 50.2 with its dome on. The dome off xhp70 might be a tad smaller.

The left pic usually has a very noticable donut hole, but I’m using a trick to reduce it. I moved the emitter slightly off center to the reflector. It doesn’t eliminate it, but it helps reduce it.

One other difference between the xhp70 and the xhp50.2 is that the 50.2 gets noticable hotter. I haven’t had a chance to take tail cap amp readings, but I’m thinking it’s higher than the 70. I’m using an FX-30 driver with a wire over the sense resistors and liitokala batteries.

Overall brightness between the sliced dome CW 70 and stock 5700k 50.2 is pretty close. I don’t have my meter right now, but i can tell the 50.2 is a bit weaker.

I was quite let down with the reduced lumens after cutting the dome off my xhp70. I’m probably going to go back to a domed xhp70. I might wait for a 70.2, but which ever I choose, I want a 6500k version.

About color, I’ve had the xhp70 N4 1A and N4 1C, both were about the same to my eyes. A bit on the nuetral side to me. I don’t care for it. I’d rather have a cooler white. This new xhp50.2 in 5700k is noticably more yellow. The corona is definitely more yellow, but the hotspot is only slightly more yellow. Combined together they create a more yellow overall looking beam compared to the CW xhp70.

I hope this helps someone.

At the beginning of this night I had fully charged batteries and my light got hot at just 20 to 30 seconds, which is definetly faster than the xhp70. After using it throughout the night I noticed it wasn’t getting hot anymore.

I did a timed run to 5 minutes and it only got a little warm. I checked the battery voltage and it’s down to 3.6v. Normally I can go 3 full nights with the xhp70 and end up with 3.7 or so volts, but this emitter sucker them down in one night. Wow! I’m not liking this new 50.2 at all. The 70.2 will probably be the same.

I’m gonna keep using it a little while longer to verify my initial impressions then I’ll probably skip the 70.2 and go back to the regular xhp70.

Djozz, how do you think the lower voltage of the xhp50.2 would do with a boost driver?

I might pair it up with Kaidomains H1-A 3 amp boost driver.

I think it will be a pretty good match, I have one on order for that reason, and khas akready did a mod with that combination (in an EE X7). The lineair drivers that burn up overvoltage are getting pretty inefficient with these new leds, a boost or buck driver may just be the better choice nowadays.

There is a very simple fix to the battery life, simply reinstall a sense resistor to limit current. The XHP50.2 has such a low Vf I would not be surprised if you are running 12A+.

If you install a sense resistor to limit power to around 6-7A the batteries will last about the same or possibly longer then the xhp70. Although most likely with a bit less lumens naturally.

I want to swap to an XHP50.2 or 70.2 in my L6 and swap back to the stock driver with the sense setup to limit it to 7-8A.

A better choice would probably be for Cree to not use this new lower voltage. :slight_smile:

I can see people hoarding the older designs already. Lol

Thanks TA. I’m charging my batteries back up right now to test the amperage.

I don’t have enough resistors to put it back to stock, but I should be getting in an order of 20 each R082, R100 and R120. Once they arrive I’ll be back in business. :slight_smile:

Heres a question for you.
If my FX-30 driver had the stock resistors (two R082’s) would the 50.2 draw more amperage that the xhp70?

I’m curious how the lower emitter voltage effects the drivers output.

The lower voltage is actually a massive improvement in everything not flashlight related and the only reason it is not a positive thing for flashlights at this point is that we don’t have drivers designed to take advantage of the low voltage.

We need some current controlled drivers, buck being the best but also impractical solution.

I have a few ideas floating around for drivers that would take advantage of the lower Vf and once we have drivers that can control the current then these will be the better option for sure.

That way you will actually be able to regulate the output instead of the output flat out dropping with the voltage.

In your case by putting the sense resistor back you will get a regulated current output instead of an unregulated voltage output. Thus the light will remain at a steady brightness until it falls out of regulation and net you better real world lighting.

With the XHP70 the Vf was so high that you had to pick between max brightness and regulation, with the XHP50.2 you get both.

In theory the sense resistor is only effected by current, so the stock 2x 0.082ohm should give about 5A regardless of the LED used. Although depending on exactly how the circuit is setup that can vary a little.

Also remember that the FX30 is a buck driver, so the tailcap current draw will vary depending on the input voltage while the output will remain more or less the same. Better batteries will actually cause a lower amp draw at full charge.

Thanks for explaining things.

It seems these new lower voltage emitters, including the 70.2 which should be out soon, are not the easy upgrade many were hoping for.