RESERVED TO:
Topics, Suggestions, Comments and Questions:
GENERAL ASPECT / DESIGN
- Change the style extruded tube with simple tube with big bore inside. With parallel connection it will allow to use 3×21700 or 4×18650 with same tube. (kiriba-ru)
- “combined/synchronized machining of head and body threads to enable precision and alignment of the 2 parts when screwing them together till the end” (kiriba-ru)
- A tube that’d hold 3× 26650 cells, and a head that can hold 3× 26mm TIRs. (Lightbringer)
- single 26650 and a Ledil cute SS triple optic in a tube light like the SP33 (ZozzV6)
- a different battery tube than the default SRK battery tube, just to distinguish themselves from the many many cheap crappy clones with the same battery tube design (Jerommel)
- Make the tub a big block knurling or big diamond shape but not the SRK (ZozzV6)
- slightly smaller (3×18650) size, it might be just that much more easily carried. (PBWilson)
- cartridge for 4 batteries (Serp)
- 3×21700 batteries (kiriba-ru)
- 4×18650 (suratyu)
- 4×18650 (Andybibbville)
- battery carrier is better than fixed tube because it is easier to mod 2S2P and replace the leds with 6V ones (ZozzV6)
- Batteries should be 18650 or 26650. 4× 18650 is a bit too fat to be easily carried. How about 2 × 26650 in parallel? An oval cylinder battery tube might be easier to hold. (Andrew_Debbie)
- Deal with the heat, have a nice and current UI (ramping?), make sure the switch works well and come in at a nice price and you’ll bound to be a winner. (PBWilson)
- The battery holder should be careful with the interior design, not to damage the battery side wall during insertion. Each slot should be well separated, and the battery will not lean over the other slot. (WestBam)
- External dimensions: Do not have a very large diameter because many people can not handle it comfortably. (WestBam)
- Head: good if the size is the same as the size of the body and not bigger. This makes it easier to carry while it is in use, taking up less space.(WestBam)
REFLECTOR / LENS / OPTICS
- 3 orange reflector and 3 smooth reflector (Jackie)
- Carclo optics with 3×4 emitters (leaftye)
- stick with fewer but intact reflectorlets (Lightbringer)
- 7 in 1 TIR (Jerommel)
- normal deep reflector for 3 LEDs (Serp)
- idea of TIR optics best, non-intersected reflectors second. (Persechini)
- Optics: Combine mixed OP and SMO reflectors. There is not much on this market yet! (WestBam)
EMITTERS AND MCPCB
- high cri nichia 219C’s (RobertB)
- 3 XHP50.2 or XHP70.2 on deeper reflectors (AlexGT)
- 3 XHP70.2 (NikolaS)
- 3 XHP35-HI (gchart)
- XP-L HI emitters (Delta_V)
- 3 XHP50 CRI90 (Lexel)
- MCPCB designed dual footprint with XHP50 thermal pad non conductive (1206 Resistor pad bridge for example) (Lexel)
- 3*XPL and 3*XPL-HI (Jackie)
- 6 Luxeon V (4000K) (WTF)
- 6 XP-L or 6 SST-40 (giorgoskok)
- XP-L HI and Nichia 219C; Copper DTP MCPCB (XXX-Man)
- Two emitters options isn’t a bad idea. XP-G2 or G3 for the budget option, and XP-L HI for the lumen hungry, maybe a third Nichia option (leaftye)
- some 4Cs/4300Ks in there, or some 4500K Nichias. Make a WW option, too. 3000K ; 4500K is low-stress on the eyes (Lightbringer)
- 7 emitters would fill the head nicely, like a flower pattern (6 in a circle with 1 in the middle) (Jerommel)
- High CRI Nichia (Jerommel)
- 3x or 4x Luxeon V and drive them hard and get close to 6000 or 8000 Lumen (Jerommel)
- warmer to neutral tint with a nice balance of flood and throw (PBWilson)
- 3 XHP50.2; 3-5mm base with a direct heatsink (Serp)
- triple design; XHP 50.2; 3 XHP 50.2 in 6-volt configuration at half of their max forward current? (Jaded)
- 3 xhp70.2 (suratyu)
- 3 xhp70.2 (Andybibbville)
- 7 (emitters) is best: uses almost all of the front surface, without dramatic cutting of the cups; no dramatic petals in the spill and decent throw (Jerommel)
- 219C, but the beam pattern and tint shift problems would be better with TIR optics. (Jerommel)
- 7x Luxeon V it is quite possible to reach 14000 Lumen (before optics) on freshly charged (high drain!) cells. (Jerommel)
- 8x Luxeon V it’s easy to get 8000 lumen, regulated for the most part of the battery charge. Turbo could be 16000 lumen. (Jerommel)
- 7 as most efficient reflector or TIR lens usage. Then don’t need to push every led so hard and they can work more efficient. (ZozzV6)
- 90+ CRI 4500K option (Andrew_Debbie)
- 7 units, all regular circular reflectors, center plus six surrounding. No reflector bastardization with great use of available area (Barkuti)
- 7 emitters (JudgeDredd)
- benefits of 8 LEDs. You can split it into 2×4 or 4×2, which is better when you have a sense resistor before the FET. (Jerommel)
- number of emitters, we have to remember it is also a matter of price, and the number 6 might have been reached as a production cost limit. If not, well, than I’m all up for a K7 instead of a K6 (Persechini)
- LED: Combine the same loads of LEDs in the SMO and OP reflectors. Eg XPL-HD is the OP reflector and XPL-HI is the SMO reflector. You get beautiful throw and a large flood in one flashlight. (WestBam)
DRIVER
- 2S/2P Boost for 3 XHP35/50 12V (Lexel)
- quality boost driver or FET driver (giorgoskok)
- driver that can supply it plenty of current and still have a great UI; just go for one of the BLF drivers. Forget all the development hassles and adopt something that works and is loved. (leaftye)
- current regulated driver (With 1S4P battery config you can supply 7× 3535 LEDs with the necessary current) (Jerommel)
- buck driver for xhp50.2@12v (Serp)
- current limited FET driver (Jerommel)
- boost driver with 7S emitter output voltage window (at least up to 28V) and 3S/4S input voltage, with 4S_18650 and 3S_21700 battery carriers. Driver can be designed to detect battery carrier type and adjust battery input voltage cut-off accordingly: down to ≈11V for 4S and down to ≈8V for 3S, for example. (Barkuti)
- implement a ludicrous beam mode by pumping ≈6A to all XP-L emitters in series. For high bin XP-Ls this would mean 14Kilolumens before stepdown (Barkuti)
USER INTERFACE
- optimize the ramping UI (comparing to the one of SP32A) making it faster from bottom to top, and more even between the lower and higher levels, without “stopping” during some luminous levels (MascaratumB)
- The non-ramping mode group should have “click + press & hold” to change modes and “fast click” to turn off (…) less likely to accidentally register a double click; I also find a short click to turn off to be more intuitive than a long press. It would also be more consistent with the ramping mode – in either group, press&hold would change brightness while a click would turn the light off. (Delta_V)
- from OFF short click last memorized mode or first mode of mode order if disabled (Lexel)
- long click from OFF Moonlight (Lexel)
- 4 clicks from ON as well lock out (Lexel)
- Option to Moonlight with down to 0.1 lumens in ramping (Lexel)
- option to enable/disable memory (Lexel)
- in modes an option to enable/disable moonlight and change mode order (Lexel)
- 3 short presses for strobe advance with long click to beacon and battery blink out, normal click to last mode (Lexel)
- enabling or disabling Memory as an option in the ramping group [If memory enabled the fast click from off would be the memorized mode. If Memory disabled the fast click from off would go to the Low (non Moonlight mode)] (vwpieces)
- use Andúril (made by ToyKeeper) (Persechini)
- more even mode-spacing in discrete-step modegroups (Lightbringer)
- fast click to turn OFF (Xoden)
- Ramping (PBWilson)
- NO STROBES (Lightbringer)
- Driver and Modes: Be Easy! Do not complicate it like BLF Q8. I got a lot of feedback on how good the description is, but the menu is still complicated. Anyone who wants to buy it is not programmed. or not want to change the final user firmware! The ramping I think is quite useful. Let the two endpoints be well interpreted and that’s all. The moonlight mode should be really low for a terrific long running time. Heat protection works reliably and independently. (WestBam)
LOW VOLTAGE PROTECTION
THERMAL STEPDOWN
- if this is not calibrated and based on MCUs spread this will be not good, some will be too high others too low, I calibrated a hundred Narsil drivers and the spread is +10°C/-15°C sometimes even 20°C (Lexel)
- get this with a external NTC on the MCPCB and raise temp to 80°C (Lexel)
- the thermal step-down should use a slow dimming down instead hard step-down’s (Lexel)
- Set high level around 2400 lumens and give it enough thermal capacity so the light can run continuously without stepping down (WTF)
OTHER ASPECTS
- add tripod mount (WTF)
- springs bypassed; brass screws holding the springs pcb (giorgoskok)
- The spring PCB at the negative side should be made with very very big traces for high current, to avoid melting if a battery is placed in reverse position (ZozzV6)
- Battery level indicator is definitely useful, the light under the button is probably the most good. I like to see a metal button on the side of the body and a lighted ring. (WestBam)
- Exterior: Have a tripod mount, when you have a high weight (over 500g) on the head and body at the end of a decent size and wall thickness hole to use a shoulder strap to it. (WestBam)
- threaded tripod hole would be good (leaftye)
QUESTIONS to the manufacturer
- How low is the Moonlight Mode? If it is not on the regular modes - that start on 100lm or 200lm - how low is it? (MascaratumB)
- I wonder what pricepoint they’re going after. If it’s going to be less expensive than a Q8 or similar 4×18650 lights that have been released recently, what options are they willing to lose? Simpler UI? Less machining? Less involved emitter setup? No spring bypass? (PBWilson)
- Since Sofirn came on the market as a budget priced light, are they going to follow that model or is this going to be in the $80-$100+ price range? If it’s going to be less expensive, drop any ideas of changing ports and box candy. A few extra o-rings should do it. (PBWilson)
-Would it be possible to make a light with Jerommel’s 7 LED configuration and have the center LED be a thrower, and the six surrounding LED’s be floods? Perhaps with independent control of them. Or is this a bad idea given the way lights get used in the real world? (Lazy-R-us)
EDIT
COMMENTS:
- A place for a light like this in the market, due to similarity with the BLF Q8, with similar UI.
- make the host available (MikeC)
- make a different kind of light like a triple led tube form 26650 light or something else because this is very close to Q8; This could be a good light because the UI is pretty good and I hope there will be a lot mass in the head but Q8 mostly filled the requirements for this type of light. (ZozzV6)
- 15k+ lumens would make this light stand out, and Nichia emitters are another way to make it special. (leaftye)
- With many 3535 LEDs (with appr. 2mm² Light Emitting Surface, like 219C and XP-G3) you can use smaller optics for the same throw, which also means they can be shorter and with 7 LEDs With a (regulated) FET driver in stead of a buck or boost driver, the driver board can be flat. So in all, they could make it a short, compact light (Jerommel)
- something a little different than the Q8, perhaps a little smaller and simplified for an even more affordable cost (PBWilson)
- make the K6 host available for sale (Jaded)
- Do not do another skyrayking. (Serp)
- Q8-inspired, C8F hybrid has a good potential (Jaded)
- dealing with heat and having an up to date UI are the most important features (PBWilson)
- Making it moddable (no glue!) would satisfy those who want to tweak it and still keep it usable and safe for those who don’t. (PBWilson)
- The difficulty with XHP LEDs is the voltage. You’ll need expensive buck or boost drivers. The .2 versions of these LEDs have some serious tint shift issues in aluminium reflectors. You need more complicated battery contact boards too. I don’t know what Sofirn wants to make, but it looks like a simple 1S4P battery set up, they’re thinking of 6 LEDs. So why not 7? Either way, you don’t have to have a buck or boost driver to drive multiple LEDs in parallel. (Jerommel)
- If they want to go for BIG output, they should use the new Luxeon V LEDs, have them run on 4.5 Amperes each, which means 31.5 Amperes in total, divided by 4 batteries is 7.875 (let’s say 8) Amperes per 18650. Luxeon V is quite good in regular aluminum reflectors. And with their low Vf (forward voltage) you can run them current regulated for a good part of the battery charge, unlike with FET DD drivers (which is why i don’t like DD drivers, they’re actually not even drivers…) (Jerommel)
- FET DD driver lights usually have high Vf LEDs like XP-L and XP-G2, because they don’t draw too much current on a freshly charged cell. Batteries and low Vf LEDs can get into trouble with DD drivers, unless you use lossy springs and thin wires, which is just a lame solution. I don’t know if Sofirn is into driver design, but it’s only a few more parts from and FET DD driver to a FET regulated driver. Probably they want a LED board with all LEDs in parallel config, only need two fat wires (16 or 18 AWG should do). (Jerommel)
- why not what looks like the acebeam x80 configuration? (about the number and disposition of the emitters) (Gerrit)
- The K6 has a better balance between complexity and flexibility. I especially like the side switch battery indicator. The lack of strobe is another plus. (Andrew_Debbie)
- Is a single emitter possible at this power level? If it is I would prefer a single emitter in a flood pattern.If not, would 5, 7 or 9 emitters be a better use of space and better beam pattern? An odd number allows a centre emitter. (Andrew_Debbie)
- ability to (factory) switch emitters with minimum to no driver reconfiguration (Barkuti)
- change the name to K7! 
(JudgeDredd)
- The Voltage drop over the sense resistor will be significant with 7 or 8 LEDs in parallel on 1 single FET. But maybe that’s just a matter of making the sense resistor 7 or 8 times smaller, but when you satrt out with 0.01 Ohm, you’ll end up with something like 0.00115 Ohm, which is possibly lower than the PCB traces and / or LED wires… (Jerommel)
- I wonder what Sofirn’s goal is for this particular light. Is it to compete with the BLF Q8 and ride the wave of its popularity? Is it going up against the Astrolux MF01 and learn from the mistakes it had? Is it an alternative to another 4×18650 light at a less expensive price point? I’m not in a hurry for an answer and maybe there’s no answer as of yet, but with all of the feedback in the Q8, MF01 and Haikelight threads, hopefully they can learn from the mistakes and produce another worthy light for us to buy. I’m hopefully optimistic based on the other Sofirn light I have. (PBWilson)
- We’ve seen a picture of the SRK, but here’s some things if you’ll be doing machining. One thing to note is that saving weight is not a good thing with this type of light. If material is machined away, it must serve a purpose. Grooving fins for heat sinking is good. Machining the LED shelf thinner is bad, very bad. I’ll just say it, leave lots of material under the LED. (leaftye)
EDIT 2
REVERSE POLARITY PROTECTION
- For polarity protection it needs plastic rings on the driver that blocks contact with negative side of battery if one reversed. (ZozzV6)
They are “into” community :sunglasses:
