Nitecore EC4 Modablity?

Flashlight Modding and DIY Parts
1

Anyone have the Nitecore EC4? Is the bezel removable? Nitecore likes to use glue / threadlocker, right? :~ Can anyone share experiences with the glue on any nitecore lights? H)

Also wondering the about driver and switch board removal. Been looking for pictures down the inside of the battery tube. Thought I saw one showing 3 Phillips screws holding the driver in. A picture in Trevilux’s great Review on ForoLinternas Translated shows the driver removed with 3 holes, maybe for screws.

I have serious trouble enjoying flashlights that I can’t disassemble and possibly mod later. :weary:

2

Bezel is easy removable, you only need te palm of your hand.

but not seems easy for mod, have you seen this pic of Lfatman? (colleague on this forum and forolinternas)

_(Lfatman Pic_)

This is his post: http://www.forolinternas.com/viewtopic.php?f=5&t=10938&start=20#p165062

3

No I didn’t find that picture. Thanks! :beer:

~ edit ~
In your review you show a picture of the driver board out of the flashlight. Did you remove it yourself? Maybe received it already separate, as sample? I was hoping you would see my thread. :bigsmile:

4

I got a 2nd one to mod and just gave up. It’s packed tight.

5

Did you get the driver pulled out at all before ending your plans?

I see that that switch board appears to have a plug in connector.
What I’m thinking is perhaps it was initially assembled by dropping in the driver with the LED mcpcb already soldered on. Led board passing through from the battery side into the head side. Through the cut out for the switch. LED board is then laid down with thermal compound and checked to be centered in the reflector. After the thermal compound has dried the reflector is removed, switch board is inserted from the head side. Switch connector plugs in.

So disassembly might be able to be done in reverse (hoping the thermal compound is not epoxy, more like silicone). Switch board removed first, led mcpcb pried up, driver + attached led board passed out the tail.

6

Hi:
I received the pieze unanodized and without electronics, after die-casting method.

7

Simply dedoming does not seem difficult.

I wonder if the reflow of a new LED would alter the sensitivity of the temperature sensor. And the same applies to the replacement of other copper PCB

8

The temperature sensor would be an NTC thermistor. You could move it over to a new led mcpcb or replace it with an equal value NTC thermistor. I have some 10k and 100k on hand, those are commonly used values for NTCs. Attach with thermal compound. Temp measurements might be a little slower.
Lfatman, by the way, thanks for coming over here to blf to comment. Every experience with the EC4 or viewpoint helps. The EC4 might be too big of a struggle… but the body looks so nice.

9

Thank you for your reply.

I have no knowledge of electronics. I guess that lower/slower temperature sensitivity could lead to flashlight overheating…

10

Made a XML2 3C led reflow successfully. I like warm white EC4 more than cool white.

It runs fine.

11

All,

I got mine (XHP50) in the mail last night. Turned it on for about 30 seconds, then decided it would need modded. haha

Anyway, I’ve got the light entirely disassembled. Wasn’t bad at all.

What you need:

Patience and good lighting
Long need nose pliers
Tweezers
Something long (~3cm+) and thin (less than 2.5mm) (like a long eyeglass screwdriver)
HOT iron
Flux and desolder braid (used after the board was out to get rid of the evil stock solder

First, desolder leads on LED board. This will be the hardest part if you don’t have a STRONG iron. I don’t know what solder they use but its a high melting point. I found that regular 60/40 doesn’t really help get it to flow, but the 62/36/2 silver bearing solder I have seemed to work very well. Be careful of the colorful wires on the switch board (along with the black temperature leads) as the insulation on those will burn when touched with the hot iron. After the MCPCB is out, gently pull the switch connector out of the driver socket. Pull by the plastic folks, not the wires.

Next, with the needle nose, I carefully squeezed and pushed down on the two plastic tabs that are near the MCPCB opposite of the switch. I pushed this side down just far enough that the tabs would stay still and not re-latch. Then, with a thin screw-driver-like-thingy, it pushed down on the opposite side of the board until it became I’m-about-too-break taught. At this point, flip over the body and reach into the battery tube with the needle nose. There are two tiny tabs on the swich side that need squeezed and pushed through a bit, like the first set I explained.

At this point it’s just about over. The driver is released and just needs to be pushed evenly and carefully down the rest of the tube.

If people are interested, I can post pictures of the parts that you are dealing with. There is an R010 sense resistor in a not so easy place to get that I’m busy working on now…

Stay tuned for either jubilation or other exotic language. :beer:

12

Thanks!
Any pictures would be great. Especially pictures of the removed boards, maybe with a ruler in the shot if you have one.

Hoping for jubilation. :beer: Take your time with your mod.

13

Nice. And thanks for the pictures. Doesn’t look fun to solder those wires in the head. Sorry, I failed to reply when you posted.

14

First steps (still in progress) are to get some baseline measurements. It seems like this driver drops out of regualtion very quickly at 4.0 volts/cell and falls steeply from there. I’m building a I vs Vf curve for my samples stock XHP50 so I can talk in amps. But input current wise Turbo maxes out my 3.2A power supply near the very top of of the Vin range.

Will keep you updated.

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.