Nitecore EC4 Modablity?

Flashlight Modding and DIY Parts
15

Well… I managed to ruin it. I did collect some meaningful data before it went poof though. Worst part about it is I don’t know what happened. Obviously one of my gazillion probes slipped and shorted something but I have no idea which one. Everything was going so well too…

Anyway driver efficiency is about 80% on Level 5 (lots of data on this some other time I’m not so bummed). Regulation is down to 7.89V in L5, LED current right at 3 amps and a Vf of 6.39V. Cell draw at it’s highest point (bottom of regulation) is 3.05A. A couple of reference points for the rest of the input/output range are: 7.6V in = 2.5A out, 7.4V in = 2A out, 7.0V in = 1.5A out, 6.3V in = .5A out. Low voltage protection, if you want to call it that, doesn’t come in until 5.25V. At this point, the LED starts flickering at ~400hz (guess) and is dim enough to stare at. It doesn’t go off until 4.73V, when there is still 7.3mA of drain. Since the light comes on in this mode, I suggest using protected cells if it’ll be carried in a bag or somewhere else that could cause accidental power-on.

Level 4 efficiency was measured at ~86%. Output current is 1280mA at 5.99Vf. Regulation is down to 6.9V when input current is ~1.25A. When voltage drops below 5.52V, the same low current high frequency flicker begins.

Level 3 efficiency also measured around 85%. Output is 430mA at 5.64V. Regulation is good down to 6.34V with input current at ~450mA. At 5.56 Volts the light abruptly turns off. Immediately prior to that the LED is seeing ~60mA. I didn’t catch a standby drain at this level.

Level 2 is interesting. Didn’t get to investigate it well before it died, but effieciency was low at ~60%. Output is 100mA at 5.32Vf. Drop out happens at 5.89V and ~95mA in. Light turns off when sweeping down at 5.56V, but won’t come back on until 6.75V. I would go as far enough to say that this mode displays some LVP characteristics.

Level 1 again has sparse data. Output was measured at 3mA at 4.98Vf. Effieciency seemed very low too. The one note I have on it puts it at <50% when at 8.4V in. That was before I was using a shunt resistor to measure current though and the error was large.

Regular standby drain is ~.35mA. Which is about a years worth on a 30Q/HG2.

16

So I got my second EC4S in the mail earlier this week. I really like the form factor and figure I’d leave one lightly modded and use my first screw-up as an experimental host.

Anyway, successful mods are as follows.

First thing about the light that made me tear into it so fast the first time was the CW emitter. I just can’t deal with them… Second one was no different. In with a 5A.

Copper board IS NOT DTP!! Also, the negative pad is connected to the thermal pad, and the trace doubles as a heat spreader. I used a dremel to grind through under the thermal pad exposing the Cu underneath (the body of this light is isolated from all circuitry, so no worries about shorting).

I also moved the thermistor directly to the switch board and eliminated the wires, therefore, thermal throttling is disabled (aka always running max current possible).

I wasn’t brave / patient enough to go after the sense resistor this time, but will eventually probably.

I also modded the positive contacts on the driver and tailcap with 5x2mm brass buttons. The stock light doesn’t make contact with unprotected cells, and it happens that they are all I use. They could still use a little more shimming to improve bump tolerance, but I’ll likely just mod/replace the springs next.

Turns out I was unhappy with the beam and tint of this XHP-50, even after the NW swap. More specifically, there was a bad yellow/green corona and a donut cross… I contemplated for a moment dropping in an 4000K MTG2, but decided against that after realizing an easy (win-win) solution to the root cause of the focus/tint issue. Using a Noctigon (.38mm thinner than stock MCPCB) plus an adhesive 3M XM-L insulator/spacer (.38mm) under the stock centering ring fixed it right up. I’ll have to test if lux improved too, but that will be another day…

Hope this helps someone else. It sure made me happy with the light! It’s a light, functional, and practical mod that fixes most of the stock lights issues.

17

Nice. :beer:
I prefer NW as well.
So the XHP-50 on the noctigon removed the yellow/green corona and a donut cross?

18

For the most part. It went from “eww” to “ehh”. LOL

It’s still a large [quad] die Cree in a deep styled reflector, so it’s not perfect. But the delta in tint across the beam is at least cut in half, my best guess is around a third of stock.

Also, I did go and measure lux. I got about a ~5% increase too, if I remember the old numbers correctly (95% confident). I don’t know if that’s due to higher output due to the noctigon or from the focusing or both. The heat transfer with the Noctigon is noticeably better than even the modded stock board, so there may be an output bump due to that in the 1-2% range.

Also, the original beam had a very noticeable bright cross in the center, especially when viewed peripherally. Now it’s an ever so slight dark cross/dot. I want to mess with it a bit more someday, but for now it’s okay. Honestly though, I think it’s at the best point for corona tint shift mitigation while maintaining an acceptable donut hole.

19

Any thoughts on replacing the XHP50 with an XHP70?

20

I know this is kind of an old thread, but I thought I’d add to it just a little.

I recently purchased the ec4 w/ xml2. I replaced the LED w/ a U4 on a 16mm Noctigon, Arctic Silver paste, and managed to stack (2) r100 resistors onto the stock r010. I got over 4A @ the LED. In case anyone is wanting to boost theirs… It takes steady hands, but it’s possible and definitely worth it IMO. It’s a very nice light. Seems to dissipate heat very efficiently.

21

Nothing wrong with bumping old threads when there is something to add. :beer:

22

I like this form factor. I wonder how it’d go with XPL-HI and driver replacement. Looks more pocketable than a C8

23

Are the batteries 2S or 2P? Can you pleasde show some pics how the driver is assembled? Is it a 17mm driver?

24

Batteries are 2S. The driver is not the typical driver you’d expect to find. It’s the shape of the battery tube, and it actually takes up 2 boards that are separated (and held together) by a plastic frame.

Imo your best bet would be to perform a current boost on the stock driver, as it seems like it would be a real pain to somehow rig another one in.

25

I’m planning to swap a dedomed 219C in to one for a semi-thrower to keep in a holster attached to my backpack. One thing I’ve been wondering is how difficult it is to remove the mechanical reverse-polarity protection Nitecore is (in)famous for, as I have lots of flat-tops and prefer to be able to use them in all my lights.

26

Thank you!

27

They’re on sale right now

Saul, your mod post above makes me wish more than ever that I still had steady hands!

28

Hey guy, can i replace the xm-l2 led with xhp50 led(6v), it’s run on 2 cell 18650, so maybe it can drive 6v led properly

29

I want to know that too

30

Does someone know the answer ?

31
32

Sorry kiriba-ru, I don’t know why but I recall having not understand your answer on the other thread…
I’ve re-read it now and, if I’m not completely stupid…you’re saying that a buck driver can drive a LED that has a Vf nearly as high as the voltage input ?
So a 2S buck driver can handle a 6V LED, such as XHP50 and Nichia 144A ?

33

Yes. In fact small difference in voltages is better (higher effincy), but I do not recommend to combine 6V leds with R mods.
If driver components were chosen for big power and you change 3v led to 6v led, you should receive same output current (and this is so for famous russian drivers from ebay).
EC4 driver gives same output power (twice less current) with 6v led, this means there is no reserve in R-L-C circuit. Combining 6V led with R mods (even thermal R mod) may cause problems.
I have cutted piece of stock board with thermal R and glued it to the head from inside. Thermal management should stay similar to the stock one.

34

I use the russian drivers found on ebay, they’re fabulous !

Anyway, back on topic I could reflow an XHP50 on the stock MCPCB, right ?