Wow - nice features you got there! Didn't know bout the dual speed PWM thing - sounds great! Batt connection flash... ? Think I've seen this happen before with some driver... Didn't think of doing hold to cycle - I do hold for one mode at a time to work back down - longer hold for strobe. I got the 1-12 blink batt check, but don't use it in my e-switch version simply because of how to access it - but I could use it simply in place of strobe of course...
Back to the L5 - emitter shelf is look'n thin - e-switch is mounted on the driver, so I would leave the stock driver in and strip parts for piggybacking a custom driver. Only 9.7 mm space above the driver PCB board to work with - not much room if you want to add some heat sinking mass under the shelf, and piggyback... The lens is 36.8 x 1.42 mm. The UCLp to match up is 37.0 x 2.25 mm. Reflector I.D. is 33.3, slightly less than a F13 which is 34.5. L5's is also shorter by 3.5mm, so not straight fwd to LEGO.
Update: UCLp result: 2.9% bump in OTF lumens, 1.3% bump in throw. I expected better - the UCLp looks gorgeous in there but the stock lens didn't look so bad either. However, I'm increasing lens thickness by ~0.8mm, so that means the bezel is extending further out by 0.8 mm from the reflector. Could it possibly be blocking more light? Maybe? The bezel has about a 3.5mm straight tube portion before the angled out section starts. Many lights have this, but I think it's more efficient for the bezel to be very low profile, like the X6, Convoy S3, cheap e-swithc XP-11, etc. Chris can only do UCLp's at 2.25 mm, while his UCL's (glass) appear to be custom made for different thickness's, but he can only do short runs at reasonable prices with UCLp.
Update #2: Found a thinner o-ring that fits. Re-measured: 710 OTF lumens and 19.2 kcd throw, so 3.3% bump in OTF lumens, and 6.9% bump in throw. I'll take it.
I'm sure there's a good deal of margins of errors here, but also, I'm using the same cell all along, and may not be giving it enough cool down period, so possibly results are even better. There are also these "variables" of lens/o-ring changes, and who knows - maybe just the act of taking it apart, putting back together could slightly change something...
The battery connect flash simply blinks once when the driver receives power, to confirm it has a connection and enough power to operate.
Hold-to-cycle will ramp up from the lowest (or current) mode to the highest, then turn off and stay that way. So, it also acts as hold-for-off, though it’s usually more of a “hold until desired level is reached”.
Battery check is 0 to 5 blinks… for ~0% (0 blinks), ~25% (1), ~50% (2), ~75% (3), ~100% (4), and > 100% (4.35V) (5 blinks). It can be accessed by quickly clicking through all modes from off back to off, or by press-holding to the first level then tapping once quickly to get back to off.
It has no strobe mode, but I’ll be making a variable-speed ramping strobe firmware later for another project.
ToyKeeper - I thought I read the battery check's we use don't work correctly with a zener mod, 2 or more cells in series. Does this firmware handle that? I know there were discussions way back, but theoretical. Dunno if it ever got resolved/published.
Nope, I haven’t even tried to make it work with a zener mod or serial-connected cells. Since the MCU voltage is artificially limited, is that even possible?
I assume it can be done if the voltage ADC pin is attached so that it bypasses the voltage limiter, but I don’t know what I’m doing on the hardware side of things. The ADC would simply give different readings, in that case, and might need a prescaler adjustment to keep the value from clipping.
Ohh - I just follow the recommended zener mod guides to assemble - pretty simple to do to a BLF DD driver or a Nanjg/Qlite. I've done a few and lots of others have done them as well, just not certain what to do about the battery level checks because I think the standard check don't work.
I also added a UCL that I have been using in my f13 clone and it fits great. Only a 3% increase in lumen with the lens. I suspect the coating is cheap as the lens doesnt look as “clear” as some of the nicer UCLp’s I have.
Numbers!
Test run on a 50a Sony vtc 26650 (Efest 35 might be slightly better as Richard measured higher numbers WITHOUT a ucl)
1512 at turn on
1401 at 30 sec (Richard measured 1575@ 30 sec on an efest 26650 in his lightbox)
739 high
99 mid
26 low
6.5 amps measured @ tail!
impressive!!
50,000 lux
this light easily makes all the f13s and similar clones obsolete at this price
well I guess if you got some f13s at 9.99 from GB like i did they still have some purpose
Anyone know of a way to get a 13mm copper XP star?
(or a larger star which can be safely filed down to 13mm)
I suppose I could reduce a Noctigon XP16 down to 13mm, but I’d have to be careful not to short the pads to the pill. I think the amount of cutting would be about like this:
It also might be a bit difficult to get the cut quite right, using only a hand file and/or dremel.
I've done this a couple of times. My thread on the Tank007 E10 mod (https://budgetlightforum.com/t/-/22831) ,I used a SinkPAD, but you can do it with a Noctigon. One trick is to file the edges at the pads at an angle so the pad edge is no where near the housing it's sitting in - worked great for me. I did a similar mod to the SingFire little SF 348, pics here: http://s1054.photobucket.com/user/TomE2012/library/SF-348.
Always check for shorts of course. I used a rotary tool with the sanding drum I believe. I don't have any trouble getting things round for some reason. Mark your target with a marker - O-L style, and use that as a guide to how much material you are taking off.
The SinkPAD's were made for this with the two pad set, but Noctigons will work too, specially for 13 mm.
Make sure to file a bevel around the edge or at least in the area of the pads, this will allow the mask to separate the pads from the star and the edges of your pill.
Thanks, I think I’ll try that. I’m just improving my cnqg brass 18650 light, and there are virtually no stock components which fit in it. Fortunately, the driver is a standard attiny13a so updating the firmware was easy. Now it just needs a better emitter (and I was hoping to add copper instead of reusing the stock aluminum star).
Hey guys, sorry I have been virtually absent this past weekend. I had gotten a little behind on flashlights and custom builds, so I pretty much shut everything off for two days and got caught back up. If you sent me an e-mail/PM and didn't get a response, please send it again as it may have gotten lost in the shuffle.
The Maxtoch lights are in! I will get them up on the site sometime today. I have to say that my initial impressions are very positive. I have a $120 Olight SR52 and $150 TN32 sitting here and these lights are every bit as good. The machining, threads, and finish are top notch--almost flawless. The lens and reflector are also excellent, and the lights are lightweight considering their size. Of course we have some work to do if we want to make them really sing, but they are about as good of a starting platform as any I've seen. I think that one of these with a few tweaks will spank the expensive lights without missing out on any of the quality.
Also, batteries and LEDs are set to arrive today. I will post when they are available.
I finally took a good look at the JAX Z1, and I think it’s an ideal host for a laserlike-zoomie mod. Seems like it should do 200kcd without a lot of trouble. If you decide to carry those, count me in.
Night Spy, input is based on the single cell Li-ion 4.2V with a 3V cut off. Output is going to vary by the cell used, with a max in the 6.8A range from a Sony C5 or Efest 35A in new condition. The lowest I’ve seen is around 4.5A from Sanyo UR18650FJ cells out of a laptop battery pack.
Maybe it would be worthwhile to point folks at HKJ’s BLF17DD review? In that review HKJ goes to no great effort to explain the concept of “direct drive” and it’s implications - so reading his review will probably only help so much if a person doesn’t already know about those things. It can be difficult to communicate in an accessible way exactly how DD ends up functioning. Maybe lagman’s thread will help people better understand the concept and it’s implications. I’ve been meaning to review his writeup but haven’t done it so far :-(.
Yep, I understand the basic concepts and the limitations of various components.
What I am not seeing is the allowable maximum input voltage to the driver for series terminated emitters.
