This is my first post, and this is my first build, so any advice would be great. I just ordered the following components from mtnelectronics.com:
Host: CONVOY S2
LED: CREE XP-G3 S5 3C LED on Noctigon 16mm MCPCB
Driver: QLITE REV.A 7135*8 3.04A LED DRIVER - 17mm
Alternate Firmware: Flash guppydrv Universal
Battery: Panasonic NCR18650PF Protected 2900mAh
Others: 20mm AR Coated Glass Lens.
I am excited to get started. Hopefully it should be a pretty nice EDC light. Feel free to point out any issues/flaws you may see with what I am using, or any advice for the actual build itself.
Unno, I think an XM-L2 would be more efficient, especially at 3A. You’re dealing with an OP reflector, so it’ll be “smeared” somewhat anyway, and you’re not dealing with a thrower where a smaller chip would be a benefit.
I got an Eagle Eye with OP reflector and micro textured lens with xpl V6-3D, runs only on 1.6A CC but great light to be in my car with white and red diffusors
Correct me if I’m wrong (again, very new to this) but according to a review that was done by djozz on the XP-G3 shows approx ~1120lm at 3A. From what I’ve found, its comparable to the XM-L2 at 3A. The XP-G3 shows 3.25V going to the led at 3A, and the XM-L2 shows 3.2-3.4V to the led at 3A. These stats seem too close to see with the naked eye, and the XM-L2 was 1. Out of stock and 2 almost twice the price.
Again, I am new, so if I am completely overlooking something on this please teach me!
I don’t want to rain in your parade, but I’m not a big fan of the XP-G3 emitters. There tends to be a big “tint shift” between the hotspot and the spill. We all have different preferences, so you might not mind.
I love the S2 and own a few, but I usually use a Nichia 219C or XP-L / XP-L HI in mine.
If you haven’t already noticed with the S2’s, you’ll likely need to solder the driver into the pill. Not a big deal. Some guys modify the retaining ring and use it, but I find it easier just to solder it. And don’t forget to use some thermal grease or something under the LED.
Good luck with your first build! Post an update with your progress.
No sweat. Now that you’ve broken the ice, I bet you’ll have more orders in the near future with MtnE. If the XP-G3 ends up not being your cup of tea, a new emitter is just a few clicks away
Very true. Plus it’s a $4 “oops”, not a $60-100 manufactured flashlight that I can’t change. That’s what got me started on building this one in the first place. Figure out what I like and keep upgrading from there.
I just try to not stress out LED chips too much, and the G-size chip would have a much higher current density than an L-size chip.
Don’t recall what’s the maximum “safe” current for each. Ie, the XM-L2 might be listed as max current = 3A (just about redline), but the XP-G3 maybe 2.5A (definitely overdriven).
I just tend to be conservative with my designs.
— later —
Okay, “by the book”, max = 2A for the XP-G3, 3A for the XM-L2. So definitely 50% overdriven for the XP-G3.
If the area of the L chip is twice that of the G chip, that means the current density of the G is 4× that of the L. Or put another way, the same wattage (heat) in ¼ the area, or 4× hotter per unit area.
So for the same exact amperage going through the G chip vs the L chip, the G would be 4× as high-strung as the L.
Point is, just because something can be done, doesn’t mean it should be done. Unless you have a setup where you’re definitely going for throw (like in a C8 or L2), the bigger chip wouldn’t make much of a difference in the final beam (smaller OP reflector), but would definitely be cozier. Less heat, less stress, better lumen-maintenance over time.
You want a thrower, put an XP-E2 on a solid Cu block with lots of heatsinking and run it ’til it’s screaming, and you’ll have an überthrower, albeit maybe short-lived.
For an S2+, again, me personally, I’d stick with an L-sized chip if running at/above 3A.
i always used xml2’s on my s2 and have used the quite too but recently went to the single side drivers because i got tired of not being able to use the retainer ring without extensive widening so i had been just torching the pill and presoldering it, then setting the qlite on it and soldering it to the presoldered part. also i always and still do cut two notches with a dremel in the pill to get a better hold on it when screwing it into the head with snap ring pliers. i just made some s2’s for a coworker to give as gifts and the single sided driver was a much cleaner install. (i did add a few chips though). before i went to the single sided driver, i was using the 8*7135 driver from simon, which is similar to the quite but the chip orientation is better for stacking chips on the battery side
So basically the XM-L2 will run cooler and more stable at 3A? I didn’t realize the chip sizes were that much different. So the Nichia 219c or the XM-L2 would have been the best options. Will the XP-G3 run too hot in this setup?
They’ll both be dissipating about the same wattage, but the G will be pushing that through a much smaller chip. Like running 400hp from a 5L engine vs a 1.2L engine. The smaller donk might do it, but it just won’t be happy about it.
Forget 219s. They can’t handle that much current w/o frying.
At 3A your fine with any of them. They're pretty much the same wattage, pretty much the same heat.
Just build it. it'll be fine.
If you don't like the beam try sanding the spacer thinner and thinner till it looks good.
If you still don't like it, try an XPL HI next time. they do great in an S2 reflector. especially the smo one. use the thin clear 3M gaskets from RMM for the tightest hotspot
Good TIR optics i.m.o.
A good TIR overlays the beam and the spill, thus eliminating the colour shift.
A pebbled surface will help too.
But then you have a floody light, not a thrower.
I personally don't like the beam of the XPG3's. They're obviously good for something but in my limited experience with XPG3's (8 or 10 of em) I haven't seen it in a flashlight where an xpg2 or XPL/XPL HI couldn't do the same or better. It's definitely brighter and more efficient than the G2 but not the XPL. The one light I built that did do well with them was a quad using a Khatod optic but it still has some tint shift towards the edge of the spill.
In reality, we're just being picky and snobbish about whitewall beam profiles because in the real world when your using it outdoors, it's not that noticeable unless your looking for it.
Smaller chip → better throw in a reflector (more of a point-source, better surface intensity).
Smaller chips would be better as a triple or quad (divided current, eg, as a triple, ~1A each from a 3A driver), vs dumping all 3A through one chip.
One of my favorite toys is a C8 with an XP-E2 emitter (E being even smaller than a G), at about 1A. Not a pocket-rocket as far as lumens, but a nice tight hotspot that throws oh so well.
I might make a second with a 1.4A driver (ie, overdriving it somewhat), but haven’t gotten around to it. 1.4A wouldn’t be pushing it too hard, but it’d still be stressing it out a bit.
Look, lots of people here try to find the maximum number of lumens they can put out. Lumens vs current is a parabola. The more current you push, the hotter the chip gets, the less the efficiency, and that much less light increase (eg, 20% more current gets you only 10% more light). Eventually, that curve flattens out, and beyond that, with any more current you push through it, you get LESS light coming out.
Don’t expect the LED to last very long doing that. And pushing stupid-high amperage through an LED will heat up the light so hot that you won’t be able to hold it for too long. It stops being a flashlight and becomes a curling-iron instead. If you want that, hey, go crazy. Me, I want a light that’s actually useful, not some shelf-queen that sits there only for bragging rights (“I built a flashlight that vomits out 2100lm!… for all of 15sec.”).
A G3? Me, I’d back off on the current, or use an L2 instead (and XM-L2 = XP-L as far as the chip).
