Finally the XHP70.2 emerges!

On page 25 of the data sheet you can see it explained: http://www.cree.com/led-components/media/documents/ds-XHP702.pdf

The Anode is V+ and cathode is Ground.

Hey T A, thanks for the quick help!

I’m doing a total overhaul.
…but the progress is a bit like a slug in winter for now…
I got some nice TIR optic from Kiriba-ru to replace the reflector (less spill and eliminating the tint shift) and looking to replace the LED shelf with 4mm thick copper disc, but someone with a lathe will have to cut out the 1.5mm thin shelf…
Driver will be doing around 6.5 or 7 Amperes per XHP70.2
So that’s 120 Watts or more.
I’ll probably get some VTC5A batteries for it too.
Hoping to hit the 15000 Lumen. :slight_smile: (okay 14000 maybe…)

So there's a TIR optic for a XHP70/XHP70.2? Interesting... I'm thinking ~8 XHP70.2's in a BLF GT .

In the MT03, vertical space is tight. Yea, the shelf is terribly thin. It's a waste of the thick metal body - just seems out of place for the walls to be so thick and the shelf so thin, and those thick plated copper plates between the cells and driver - nonsensical.

I replaced the stripped head screws on the battery tube today - had to use a dremel cutting blade to but slits in the tops of them, then use a regular screw driver to remove them.

Also drilled out clearance for the LED wires through the shelf, eliminated all the rough edges.

The BLF SRK v3.4 HQ driver is reflowed, set up as FET+1 using the Hyperion FET. Using DEL's new C3 and C4 caps in the design to tone down spikes the 7135 takes in these 2S high amp setups.

Yeah, very odd they did that.
I guess they wanted heat management for the driver?
They could better have used bigger FETs, diodes and coils i.m.o.
But by reducing the thickness of those discs i could gain a couple of milimeters vertical space.
Also the TIRs are a little less deep than the reflectors, and there’s some space between the optics and the glass lens too.
I think i can make the modifications fit.

I was thinking to make centre a hole in the copper shelf for the wires.
I’ll be using 18AWG.
That centre hole can be large too, since the thermal path goes to the outside.
For now i have only bypassed all springs, with a double bypass for the + spring of the battery compartment, i removed the inner spring.
Anyway, i have all the parts to continue the mod now, so i should go ahead soon.

i have ordered a few xhp70.2 from digikey, and disappointly have burnt up 2x, i am coming to the conclusion that these are much more current sensitive in my setup than the xhp70, and probably different from one to another. i noticed the meter to spike pass 18A both times (narsil ramping).

first, using a s70s as a host with TA FET Narsil driver, updated with SIR404, 16awg, spring bypass, littokalas

second, same setup but with reduced 22awg (this setup works with my L6, but it only draws 16A)

i am not sure how to get these to work at +18A, any suggestions?
has anyone found the max safe current?

https://www.digikey.com/short/31d2jm

LedNLight has some XHP70.2 TIRs listed. There are more manufacturers of XHP70 compatible TIRs, but I haven’t collected the data.

Poor re-flow / heat sinking maybe? Here is my SD75 2S2P bypassed battery carrier, Narsil Lexel built T/A SIR800 pulling just over 20 amps on VTC5’s, so that’s what I run in it. No problems so far…

My VTC5A’s turned the 70.2 instantly blue at 22-23 amps IIRC? :open_mouth:

I measured between the driver lead and the LED.

At extreme currents like 18A everything has to be perfect to keep the LED alive. That is a truly insane amount of power for such a small comportment. Over 120W worth of power to be exact!

To be honest I am impressed that they work at all at those currents. I would never expect it to.

Dropping the current down to 10-12A is what I would personally do. Above that you will reach a point of exponentially more heat and less extra lumens as power increases.

I measured 17.9A at the tail with a clamp meter on my L6 and it's surviving so far. The cells will drop pretty quick at this rate, and the heat is a problem of course. On LK cells w/solder top, it started at 9,100 lumens, dropped to 8,230 in 30 secs.

you guys are nuts lol.

I have over (12) XHP70/70.2 lights in my keeper (never to sell) collection, and this little UTorch UT02 is one of my favorites!

Machined out the head for a 32mm Maxtoch DTP board, 70.2 P2-1A Lexel built T/A Narsil driver, with reasonable out put 9amps at turn on and just over 8 amps around 30-40 seconds, or how ever long it took to try and grab the shot of the clamp meter in 3 attempts… :smiley:

i am not sure how to know if the reflow is any good, maybe when i “bop” the led it created voids in the solder. i will try again. without “bopping” the led. any reflow tips?

maybe after this i will learn to accept less than insane maximum :slight_smile:

(Had to fix that for you :wink: ^)

I BOP them too, maybe too little solder and BOPPING does no good? I usually have squirt outs at 4 points after BOPPING?

Did the LED go blue before Poofing? Or was it an instant Poof?

If it was a long slow burn out (meaning not an instant Poof) maybe thermal contact was insufficient?

Please don’t get discouraged, It’s a bummer, yes, but 8-9,000lm out of these 70.2’s after all, is quite something to behold!! :wink: :+1:

when i bop the led some solder did squirt out, i then push the led a tad to see if it will pull itself back onto the pad. i wonder instead of bopping to squirt, i would bop and hold till the led cools down for better contact.

the led did go blue before “poofing”, long enough for me to realize “oh sh!t” that is not the color i want.

i wish Narsil have a ramp speed setting, current setting is too fast for me to stop it from going full power, a much slower speed would allow me to stop once i see some blue tint.

Be careful you are not overheating the LED when reflowing as well. It is easy to do if you are not using a temperature controlled heating source.

good thoughts, going to make use of the FLIR one for the next reflow :slight_smile:

Could you fit more than 8?

Dunno, only a wild guess here because of course the GT is the widest head light around.

I noticed with the 3 XHP70.2 P2 3B's in the MT03, I don't notice the yellow ring around the halo. Not sure it made a difference, but I scraped off the yellow on the edges of the LED. Pretty sure my MT07 and L6 have the yellow ring.

I'll have to try it on the L6 or MT07 and see if it makes a difference.

You think that might reduce the yellow around the main beam? Let us know of your findings, I was wondering if a sliced dome close to the surface of the die might reduce that too…