TK’s M43 seems broken. None of my four M43 lights have this issue. Turbo is accessible until resting voltage of four 30Q cells is 3.45V, and with four GA cells at 3.50V. Or the light blinks when it’s stepping down because it gets too hot. With turbo this is around the 1 minute and 30 seconds mark.
My M43 with XP-G2 emits 7291 lumens in turbo and withstands 180 seconds. The temperature it reaches then is 75 degrees Celsius. Evidently, TK has a bad M43
It doesn’t seem to be just mine. It sounds like this type of issue is somewhat common. It has a complex design which sometimes works and sometimes does not. People have complained about the UI, but it has issues beyond that… like a constant current regulator which often fails to keep the current constant… even at low levels. A thermal regulation algorithm which oscillates and has chaotic behavior. A battery protection algorithm which sometimes works and sometimes makes turbo fail even when the batteries can handle it. Most of this seems to be caused by the firmware.
Making it less complicated would probably help improve consistency. Giving it some easily-accessed calibration functions would probably help improve consistency too, since the MCU is known to vary quite a bit from one chip to the next.
The hardware Hank produced is very nice… Hank always does a great job making his products high quality. It seems like it just needs a few things fixed in firmware.
Interesting. I recall you wrote about your M43 outputting less than specification over, what two years ago or so? That was the first and only time I read about someone with issues with the M43. And now the second person with problems. How often have you read or heard about issues, or did you hear this from Hank?
If you have Domed LEDs your light should be closer to komeko s light —My light is Dedomed —with the 507 optics I use to reach around 6800 lumens with VTC5 cells at turn on —- since then I changed to 511 optic polished —it reaches on average 6350 lumens at turn on
That sounds about right. I also have the S4 3D Meteor. I don’t have four VTC5A cells though, so I haven’t been able to get mine up that high. I checked just now with four 30Q cells at 4.07V, and got 5775 lm. Or with four LG cells at 3.99V, it got 4850 lm.
I finally have a decent measuring device calibrated with maukka lights, and I checked the calibration with three reference lights right before the test, so these numbers are probably reasonably accurate. I’m still charging up the 30Q cells for another test though, to see how they perform when totally full. I think I may be able to get it over 6000 lm.
Out of curiosity, I took one of those LG cells and put it into my D18. The D18 is at a bit of a disadvantage since it has less-efficient high-CRI emitters, and it was running with only one cell, and that cell was at a slightly lower voltage after the previous test — 3.94V. But it measured 6040 lm on one cell, compared to the Meteor’s 4950 lm with all four cells.
Edit: After charging the 4x30Q cells, I tested again… but output actually went down, not up. With full cells, it measured 5423 lm. Earlier, it got 5775 lm with the same cells at a lower voltage.
That just means I was holding the light and moving it around during the test to emulate a more IRL usage.
Sometimes people will use a fan to cool the light in a stationary position or leave it standing with no contact during tests.
Probably better to ask here. Can the m43 be reflashed to andruil or other firmware? I’d like to take the complexity out of the m43 and go for a more simplified UI.