I know you were using this as an example, but not sure if you are serious bout this. I was wondering about Maukka's poor efficiency measurement, whether it is equal to a PWM'ed FET driver, like the A6/ Bistro/Narsil based FET+1 drivers. Of course the modes operating on the FET and less than 100% are inefficient - goes with the territory, just wonder if the S2 driver is the same or worse. I suspect it was the same - doing the same thing we are doing, PWM'ing a FET so efficiency is not good, unlike a buck, boost or true PWM-less current controlled driver.
Of course down side of buck and boost drivers is they are complex and expensive, down side of current controlled is unable to do high amps as easy - again, complex, can be pricey, and amp limited.
Entirely serious. Hereās Maukkaās graph of the S42 at 443 lumens running on a VTC6 (3000 mAh):
And mine of an Astrolux SS running at what I estimate to be 414 lumens with my smartphone and integrating shoebox on an HG2 (3000 mAh):
215 minutes is nearly twice as long as 112 minutes at nearly the same output with a similar shaped graph. This comparison isnāt quite apples to apples: the XP-L HI may be a little more efficient than the 219C, but one emitter is a little less efficient than three. The VTC6 tends to have slightly more capacity than the HG2. A bit of difference is to be expected here, but not twice the runtime.
The ~1500 lumen output maukka measured from the S42 on a VTC6 also falls well short of what Iād expect from a 219C FET quad. My triple, with D240 bin 219Cs and a MTN17DDm is over 2500lm@30s on an HG2.
Yea, the 219C 90 CRI is a low bin 219C. If the XPL HI is a high bin of V2 or V3, then there's a pretty big difference in efficiency. There's a price to pay for high CRI, and the price is measured lumens. The new gen CREE's especially are like this. You can now get at top, or near top bin CREE LED's in neutral tints, but at low CRI, 70 or lower. To get high CRI, you drop a good amount of bins.
To get 2,500 lumens from a triple with D240 bin 219C's, it's gotta be crank'n a lot of amps, I'd guess 13A-14A (maybe more) based on TA's 219C D240 tests here: https://budgetlightforum.com/t/-/43426, taking TIR optics loss's into account, which is considerable.
Still, 1,500 lumens from a quad taking 10.6A doesn't make sense either - there's some major loss's goin on somewhere.
~15A sounds about right to me for a 219C triple with good bypasses. I donāt have a clamp meter, so I can only estimate based on light and sanity check that based on heat. For a quad to pull so much less, and then not make the expected amount of light from that, somethingās definitely off.
Your logic is sound, but he knowās what heās doing its more the convenience factor, plus it was coming with a new eFest 18350ā¦ at least it would have been if it would fit in the lightā¦ :person_facepalming:
The light looks very nice. But not as bright as the S41. The s42 has a very blue tint on the XPG3. The S42 is more floody than the S41. The S41 has a nice hotspot. Im thinking of changing the led from the S41 into the S42 to get a better tint.
Also the USB cover does not stay still. And yes I ruined the rear end of my Keeppower 18650 the cell is 34.3mm)
You care about tint and you went with Cree emitters? Were you just prioritizing brightness?
You could get these and get nicer tint with no visually-detectable effect on brightness. You can order them mounted on this MCPCB if you donāt want to reflow yourself. Personally, Iād go with higher CRI over a higher flux bin in the same emitter every time; human perception of brightness is not linear, and a 25% difference is right on the edge of perceptible.
High CRI actually helps MORE than a few extra lumens in helping out with seeing details. Once youāve tried it in reality and compared with low CRI, I guarantee that you will prioritize high CRI over lumens every time.
I agree. It would take a 100% difference in lumens, a specialized application or some other characteristic different between models before Iād consider a low-CRI model over high-CRI. Partial exception for the 80 CRI 219C in 5000K, which has a nicer looking tint than the 90 CRI 5000K from the last group buy; I have yet to see the 90 CRI 5000K in the S42.
The first 3 is bout right, the last 3 I'm not personally sure of - still not sure if the efficiency is that that bad or not - UI has probs, but dunno bout the worse ever - it's usable, and can't vouch for the artifacts. My S41 was an early one, and the MCPCB is definitely different than later S41's and even the S42.
Of course bout the head design handling the heat, there's basically no tube light that has a head design that can handle high amp heat - it's all a matter of what you want and what risks you are willing to take. Perfect temp regulation is maybe what you want because it will give you turbo output, though briefly.
Based on Maukkaās test, we can be sure itās a 219C, 5000K, R9050 (90+ CRI, 50+ R9). That doesnāt tell us what the exact tint bin is; itās probably sm507, as tighter binning costs more and this light is very inexpensive with four of them, nor the flux bin, but itās probably not the highest available (that being D240, last I checked).