High 1
High 2
Turbo
Interesting. I’m not seeing a comparable illumination to other lights that I’ve built that had similar numbers. Is it because of the 4 triple optics? Normally a triple doesn’t compare well to other lights with similar output due to the much wider beam angle. Is that what’s going on here? Everything is saturated from all the rains so the darker materials may be absorbing more light? I don’t know.
Tint looks good - At least in the pics I don’t see the usual DD Green.
Thanks,
-Chuck
Awesome news, thanks Dale!
Not sure if I missed it, but were you able to measure the LED current in your M43?
You didn’t miss it, just got done…
At the negative lead going to the mcpcb, I see 8.42A in Turbo. Forward Voltage is 10.03V in Turbo. So it’s doing 84.4526 Watts.
Now, what do these things mean? I measured 4.06V with the light off, this is the reading taken using the positive and negative leads On the mcpcb. Light on, in Low 2, shows 7.80Vf, in High 1 shows 8.25V, and then in Turbo it shows 10.03V.
I don’t understand it, I just read the meter on these things.
Edit: y’all show me where I”m wrong, but I see this as 2.81A per emitter at 3.34Vf each. Is this correct? We are running 3 XP-L’s in series, 4 sets of these in parallel. How all that inter-relates is confusing to me. But if you read the charts and do the comparisons, 2.81A at 3.34Vf is pretty much equivalent to the 1000 lumens per emitter the light box is showing. Yes? Match’s charts show an XM-L2 at 1070-1080 lumens correlated to these numbers, so everything looks pretty much accounted for.
It just gets crazy when you see the size of the light and 12 emitters doing this. 
Edit II: Oh, and it shows 57.5Kcd for 479.58M extrapolated.
Thanks for the measurements.
It seems like the M43 is pushing more power into the LEDs since the body heated up faster.
I wonder if that’s because XP-G2 has a higher forward voltage than XP-L. I’ve been looking for djozz’s current vs Vf plots but couldn’t find them. I think I trust them more than Cree’s own conservative tests.
I used Match’s charts, since the XP-L is an XM-L2 die. They’re not exactly the same but it’s close enough for government work.
Hmm the XM-L2/XP-L does seems to have a slightly higher forward voltage than the XP-G2.
Assuming the emitters are in a 3S4P (total Vf of 10.03V in turbo) configuration, the emitters are still “only” getting 8.42A/4 = 2.1A per emitter, which is pretty light for an XP-L. However, we are reaching the max reasonable heat generated for such a small body, so I don’t think there’s too much sense in pushing the M43 farther.
I wonder what the sense resistor values are for the XP-L edition and the XP-G2 edition.
Um, Ryan? If we have 3S 4P, wouldn’t the 8.42A be divided by the 3 in series? Each of the the 4 triples is getting the same 8.42A, divided into the 3 emitters in each chain, for 2.81A each, which is right inline with the 1000 lumens each emitter is putting out. Right?
And in that line of thought, an XP-G2 at 2.81A will only make what, 550 to 600 lumens? So this explains the huge difference in output as compared to the smaller dies. Doesn’t it?
One Sipik SK68 makes for a neat little beam, what about a dozen rubber-banded together?
Edit: Sorry, the brain wormhole at work…
Edit II: I took the 12,523 lumens, divided it by the 84 Watts, for 149 lumens per watt, out the front. I think that’s pretty dang decent, isn’t it?
Djozz just tested a new S4 xpg2 at over 900 lumens at 2.8amps.
That’s interesting, especially when 12 XP-G2 S4 2B’s in the same light made 7690 lumens (from memory, which is flawed) as compared to 12,523. Same driver, same cells, only thing different is the emitters.
What are we to make of that? In this particular instance it’s 642 lumens per emitter vs 1044, a significant difference. Again, I don’t understand the in’s and out’s, I just read the meter.
That’s also bare emitter output I believe, so no optic/lens losses
Hmm here’s what I sketched up previously for another member:
Since there are four emitter banks in parallel, the 8.42A is split among the four while the voltage is the same across the banks, and each emitter is seeing 10.0V/3 = 3.33V.
If it were the other way around, each emitter would see 2.81A at 10.0V/4 = 2.5V.
Is the camera settings the same? Seem like your triple XP-L X6 shot is righter than M43 Turbo. Building and surround pavement are dry, doubt it has any impact to the pic.
[quote=DBCstm] Interesting. I'm not seeing a comparable illumination to other lights that I've built that had similar numbers. Is it because of the 4 triple optics? Normally a triple doesn't compare well to other lights with similar output due to the much wider beam angle. Is that what's going on here? Everything is saturated from all the rains so the darker materials may be absorbing more light? I don't know. [/quote]
I see what you’re saying, except for one thing.
An Cree XP-L emitter can’t make 1000 lumens at 2.1A. My Texas Poker neck light is a XP-L V6 3D, running 3.32A for 1017 lumens, no optics but a reflector. The math won’t make an XP-L perform that high at that low an amperage. Be nice though if it would.
That’s true, but if the emitters are getting 2.81A, it would mean that the Vf is 10.0V/4 = 2.5V, which doesn’t sound right. I haven’t seen a M43 myself, but I think the wiring is 3S4P as opposed to 4S3P given the total Vf of 10.0V.
Hm…
Wonder if should try swapping the emitters in my M43.
I currently have a 1D dedomed. Tint has that aweful greenish cast at low power levels. But at max power at turn-on my luxmeter says it outputs 93,000 lux.
Dale, this is how you calculate
8.42A output current measured. Now there are 4 sets of 3LEDs in series, each of the 4 sets divide the current between them, so each triple set gets 2.1A. Each of the dividing triple has 3 LEDs in series (as the name implies) which means each gets the same current, meaning 2.1A.
So let's imagine it would 4 sets with 4 LEDs in series, then the current it is still 2.1A just that the Vf would be 13.37V opposed to measured 10.03V. Just an example of course.
Now you didn't say which XP-L HI you have, but as far as I see at Mouser they only have weak U4 and U5 bins, not the V2 and V3 top bins. The XP-L U5 is rated 720lumen at 25C (which is a very low temperature and unrealistic), rest of the calculations are easy to make from that 720lm value.
Thanks Hikelite. Yes, the set-up is indeed 4 sets of 3 series. The positive coming in to the mcpcb is in the middle, shared by the surrounding 4 sets of 3. The negative coming in is on the outer ring, the top loaded copper ring is a boost to the trace so that it surrounds the 4 sets of 3 with ample current carrying capability. I had one of the new emitters not get connection and it of course killed that string of 3, the other 3 sets worked fine. Had to take it apart, reflow it again and make sure that set of 3 were all connected properly, seated with a little nudge.
So yes, the incoming to the board is 8.42A at 10.03V.
I just didn’t see how these lower binned emitters could make the sheer output I’m seeing on 2.1A. Does’t compute. I know the beamshot doesn’t represent well, and I’m figuring that is due to the small 10mm TIR’s in the Optics. By it’s very nature, the Carclo 10507 is more of a floody beast.
That comparison with my 2 photo’s is misleading due to this, the optics in the X6 are massive by comparison, 35mm vs 20mm, with TIR’s individually measuring 17mm across instead of 10 and some 12-13mm in height instead of 6.
I know that in the house, the M43 is an intense monster now, white light making the room almost sparkle in it’s intensity. Very similar to the effect my BTU Shocker Triple XHP-70 does at a measured 14,455 lumens. Each of those 3 emitters is covered with an individual TIR measuring 35mm x 17mm tall. Another huge step up in size for the TIR department. (Yes, I used optical adhesive to glue a TIR into each reflector cup to eliminate the donut hole the XHP-70 makes)
Wish I understood what was going on here. It must be greatly to do with the wider beam profile of the 4 sets of small triples, I don’t know why else it would seem dim by comparison. And these are 4 brand new Carclo’s that I just pulled out of a tape of 10. (They come in a plastic tape much like the emitter does, 2 wide in this case)
Looking at output charts, 2.1A should be approx 850 lumens, with 12 of these making an estimated 10,200 lumens. At ~20 seconds, the light is indeed making 10,382 in my light box. The start value is much higher though, when the cells are still fresh and not taking a beating from the high load, 12,113 on the Efest cells, 12,524 on the LG HE4 cells that are new. So is this all within reason? Running the light for a little while and letting it warm up pretty good, the shift between the measured 7142 in High level 2 steps up pretty nice to the Turbo, but it’s not a massive gain. 7142-10300 or so would indeed be such a step to the eyes, right?
I just don’t know why the beamshot looked so comparatively weak, unless it’s the combination of a wetter, darker surface and the floodier Quad Triples.
Mouser has only U5 and U4:
Ratings at 2.1A and 25Celsius (way unrealistic)
U4 - 682lumen
U5 - 722lumen
Then you got loses in the optics and glass, cut 10-15% there easily.
There higher bins are U6 V2 V3 which are not sold by Mouser