I thought about the solder but did not want to chance killing the magnet, I like it.
[♛ FreemeGB] Fireflies PL47 Gen II 4*XP-L/ Nichia/ SST20 Hi CRI 21700 Right Angle Flashlight - ENDED
MWS, I can’t tell from their pictures but I get the impression they just put their black wrap over the existing cell. At the corners it seems almost “padded” as though this is the case (a factory wrapper inside the Sofirn wrapper) If not, I would strip the wrapper and leave the bare metal can, then wrap that in the ultra thin Kapton tape to prevent it grounding inside the tube. I’d use the cell, wouldn’t be a waste of money. Ultimately I’d keep an eye on it of course and when a proven better cell came along I’d retire the Sofirn cell. But I’d get some use out of it first.
Me? If that still wouldn’t work I’d bore the tube.
Yeah Yeah Yeah, everybody doesn’t have a lathe. But you don’t have to have a lathe, a wooden dowel and sandpaper can get it done. Where there’s a will, there’s a way. 
(Old Lumens would be proud! 
)
actually I believe he means that the washer makes contact to the tube sooner and he doesn’t have to compress the spring near as much?
the washer just needs to clear the spring.
Shirnask, it’s always something isn’t it? lol
I think at this point we should just wait for Fireflies input. Hopefully he’ll solve what still appears to be solvable.
I agree.
I suspect that the copper washer mod is only working because the washer is contacting the -Ve of the cell and the outer trace of the tail PCB, acting a bit like a retaining ring by providing a continuity path from end of the tube to the tail PCB, that isn’t there in the original design, and bypassing the tail spring.
Meanwhile the tailcap will be hanging on by a few threads less (depending on thickness of the washer), and the o-ring in a different position than designed.
You could also try something different such as a circle of solid copper wire at the head, to effectively lengthen the tube. Anneal it first to dead-soft (just get it red hot, and pickle off the scale in some acid, or polish it). If the o-ring ends up displaced from the sealing surface in the head (I didn’t see any such thing on the photos, just seemed like rough threading all the way), PTFE tape around the threads might solve that.
Tom Tom you are mistaken, the tube must make contact with the copper washer before there is continuity and can be disconnected by less than a quarter turn.
He’s only partially mistaken, the spring pushes the parts away from each other and breaks the contact. The better way to do it is to allow the cell to pass through the washer, or use a copper wire like that which is in the center of coaxial cable. This will effectively make the battery tube longer, like Tom Tom said, and not interfere with the cell or the o-ring placements.
Believe it or not, in the 500+ lights I’ve modded this HAS come up before… 
Which ever end I add copper wire to to gain the 1.5 to 2 mm extra length will necessarily displace the o-ring on that end of the tube by the same amount. Also the head is quite deep and a pain to solder. Also it took less than 1 minute to pull a washer out of the drawer and drop it in the tail cap for an effective solution. I can shape and solder a piece of wire on the driver but that’s a lot of effort to gain very little
Dale, I know this is just small change for you but it is a no effort fix for those less capable
Shipping always takes longer this time of year, especially when coming from China + not DHL.
I got notice for tomorrow delivery USPS. Sst20/XPL-hi 5k.
Thanks for the pic Shirnask, I don’t have mine yet so I couldn’t know just what you were facing there. Sure, the washer works so use it, when that cell gets tired you’ll know to get a different kind. (I’d forget probably, but that’s just me)
It is all good. You don’t have to solder the copper wire option into the head, no more than your copper washer in the tail would benefit from that (as-if, even possible.)
Try it out yourself, it’s not permanent.
I’m not sure that Dale has his yet (special delivery ? 
, nor do I, so this is all just speculation for now.
It’s early days yet, and the majority of those who are quietly contented with these things that are slowly trickling out, are unlikely to comment. It’s only the discontented ones who are likely to be vocal. Usually with good reason.
And the people like me who like to look at the detailed design, and review, and make constructive suggestions for possible improvement, before we actually have them, should be treated with a pinch of salt.
Can someone post a picture of the PL47 next to an Emisar D4 with 18650 tube?
Mine is in Texas, somewhere, it’s just on a sightseeing tour or something. Last seen in Lubbock, wonder what the next mystery city will be? lol
I am using thrunite th30 headstrap on my pl47. It works great.
Nice find Newlumen! Thanks for letting us know!
Yesterday I recieved my two PL47 ordered on day 1 of the GB. Tracking number never worked and I was just about to inquire with FF why tracking number cant be found.
The PL47 is much smaller than I thought. It is shorter than about every single 18650 lights on the market.
I measured light output at turn on (1s) using the Texas Ace lumen tube calibrated with Maukka lights and here are the results:
Black PL47 with XP-L HI 3A V2:
4,405 with unlabeled Samsung 30T
4,310 with Vapcell 40T
3,879 with OSO 40T
4,069 with Vapcell VTC6A
3,707 with QSO VTC6A
3,103 with Vapcel NCR21700GA button top (not standard recessed top so much higher resistance)
3,612 with Efest LG M50
It is very strange because the second day I tested it, the 30T won’t exceed 4,138 lumens and both the Vapcell 40T and VTC6A measures about 3,880 lumens. I recharged them to the max several times and same result. I wonder is it because springs don’t compress as hard after several uses or the emitters lost some performance after the initial high currents tests.
Champagne yellow PL47 with 219B sw45k 9080:
1,810 with NCR21700GA button top (not standard recessed top so much higher resistance)
1,948 with Efest LG M50
1,379 with Vapcell VTC6A (drops even quicker than with the other two batteries. scary!)
Fortunately PL47 219B still works fine after that few seconds of turbo on the VTC6A and output didn’t seem to reduce after retesting with the LG M50.
I need to find a way to charge the recessed top Vapcell NCR21700GA so I can test it against the Efest LG M50. Vapcell_Dennis said the NCR21700A button top version has a lot more resistance than recessed top.
For comparison purpose, the D4S XP-L HI 5D V2 measure 3,577 lumens using Shockli 26650 and the D4 XP-L HI 5D V2 measures 3,310 lumens using VTC5D.
Tint on the PL-47 XP-L HI 3A V2 version looks better (slightly rosier) than on the ROT-66 XP-L HI 3A V3. Fireflies was not able to source more of the 3A V3 so the PL47 and E07 will be using the 3A V2, so will be 10% dimmer but slightly better tint. Not sure if the current batch of ROT66 will be V2 or still using the V3. The 219B 9080 looks just as pleasant as on the ROT66 219B 9080 if not a bit rosier.
Without batteries, both the black XP-L HI and champagne yellow 219B weigh exactly the same at 78.8g (2.78oz)
The following 21700 cells slides easily into both PL47:
unlabeled Samsung 30T
Vapcell 40T
QSO 40T
Vapcell VTC6A
QSO VTC6A
Vapcell NCR21700GA button top
Vapcell NCR21700GA recessed top
Efest LG M50
Only the Acebeam Protected cell is too long to fit but slides in without problem. I dont think the FET driver would work with protected cells anyways so no point for FF to size the light for something that is not compatible, although I doubt they even have a protected 21700 to test with.
The included headband is crap. It fits the body loosely and I would be afraid to force the PL47 head into the holder as it might scratch off the anodizing. However, the PL47 fits perfectly in the Olight H2R Nova headband and Armytek hard plastic headband. These are my favorite headbands because they are so convenient to use. The Armytek one is by far the easiest to attach/detach headlamp but might not be as secure as the Olight H2R Nova headband. So unless you plan to do some really rigorous physical activities I recommend the Armytek headband because you can easily remove or insert headlamp while wearing the headband.
The optic seems a bit dirty but it doesn’t affect the beam at all. I wouldn’t have noticed it if this issue wasn’t posted on this thread. I have a strong feeling these lights are hand assembled by Jack himself after we order them in order to accommodate so many possible option combinations. He probably doesn’t wear gloves and not working in a dust free environment (Jack please wear gloves next time and wipe the optic with a cloth if it is dirty because the PL47 head is unfortunately press fit so we cannot remove it to clean). I’m guessing he is super busy lately since there are so many PL47 and E07 orders he has to fulfill. He hasn’t responded to my repeated questions regarding the E07 that I said I need to know before I order a couple from him. I’m literally waving money in his face with zero response.
Overall, I am very pleased with the PL47 considering how it is cheaper than anything in its performance range. Actually there’s nothing even close besides the Emisar offerings. The Aux LEDs are so pretty I would buy a bunch for gifting if only it had built-in usb recharge.
Thx skv, totally useful!
Nicely done.