TK's Emisar D4 review

I agree with radellaf . I find myself carrying the D4 except when working. I don’t want to scratch it up in my pocket with my keys. It is just so much fun and practical at the same time. If I am outside by the firepit, I can light em up with turbo or just use it to find the beer. I would prefer USB charging for a travel light though, just for convenience. Might get a second one in the future or the D4S. Mine is the XP D2 S4 3D version.

is there any mods required for the clip ? Anyone using other clips?

Yes. A ‘captive’ clip like this won’t allow the tailcap to thread fully onto the battery tube. This creates a gap between the end of the tube and the tailcap’s board, so there’s no electrical contact.

A washer is needed to fill this gap. Options include a wave washer, copper crush washer, or even a piece of solid copper wire or paperclip formed into a circle.

I use Solarforce clips on my Emisars. They’re a bit bulkier and squared-off than the Overready clip, but less expensive.

I made wave washers out of some flat brass washers I found in my junk drawer. I used two pairs of needle-nosed pliers to create the waves, then sanded the peaks to increase the contact area.

The Solarforce clip hole is slightly too small to fit over the battery tube’s threads. I used a Dremel with a sanding drum to open it up slightly. I also ground off the lanyard ring. I removed the o-ring, installed the clip, then re-installed the o-ring and tailcap.

emarkd had to Dremel his Overready clip to fit, too.

I made my washers slightly thicker than the clips’ mounting rings, so the clips can be rotated.

I don’t mind the moon memo at all, but yes, I noticed the D4S had some user-requested changes in its simplified version of your new firmware. I’m amazed at how far a one-button interface can be pushed. Certain other products seem to resort to bluetooth remote control long before this, but then they’re over $100, mostly rubber not metal, and have much bigger markets.

I am interested, but not too excited about what it takes to remove the driver and buy enough bits to grip the chip. I’m an EE and electronics hobbyist so I certainly could do it and won’t rule it out, but it’d be a relative lot of stuff to get that as far as I know I’d never use for anything else. And, I don’t think I’d get it back together quite as nice with the paste and everything. I think it’s that I only like taking apart stuff that is an extra one, something I don’t care about breaking, or don’t intend to ever put together. I’m not into modifying one unit of something I like. The D4S with the exposed pads is a more likely target, though. The lightning and candle features I could add are definitely tempting.

If it shows where my interests lie, I was willing to spend about the same money as the flashing tools (I think, ~$30?) for the clamp meter. I figure I will have other uses for that, though. Maybe. Wish it would measure AC power without having to clamp only one wire in a power cord.

I wrote a few hundred lines of C firmware for a Cypress PSOC based USB light meter, but when I look at the Arduino code for the instrument at my new job, it’s pretty intimidating. I see it call “avrdude” at some point and thought of this thread. I still like 1990s BASIC with line numbers. Did everything I wanted for much more complicated programs than these, and never gave me bugs from unintended variable type casts or overflows.

Anyway, the flashlight isn’t for tactical or practical purposes, and I have lots of light-up toys, so I’m happy to have it for what it is in stock configuration. If I ever look for a 2nd one, though, or a new version (D4X?), and expert-mode firmware is a paid option, I’d probably go for it.

I appreciate for the welcome-backs and the info folks!

Earlier today I read through 20 or so pages of the Attiny25/45/85 FW Development Thread ranging from a few pages before my last posts in that thread up to page 39. This seems to have covered most of DEL’s work on adjusting the decoupling as well as the foundations of deleting the voltage divider. I’ll do my best to catch the rest of the way up on that thread soon.

@Tom Tom - your dropbox link did not work for me :frowning: If possible I’d love a link to a post with a current circuit diagram and/or a list of current component values. Catching up on the FW development thread has helped and I probably have a good handle on the changes that have been made but it never hurts to be sure. The high quiescent current of previous designs is actually what kept me away from e-switch lights in the past. I just wasn’t comfortable with a light being dead in a handful of weeks if I failed to do a mechanical/electrical lockout.

@Dale - I agree, gotta be sensible.

@TK - Thanks :slight_smile: Rather than continue to derail your review thread I’m going to try and get my thoughts together over here: wight catchup

It’s amazing what one can do with one pixel and one button.

If you think the white version is fancy, you should see the firmware which uses a multi-color pixel. :stuck_out_tongue:
(rainbow lightsaber driver, lets the user build their own color patterns similar to how an analog synthesizer lets musicians build their own unique sounds)

Don’t worry, that train left about 4000 comments ago. :slight_smile:

I think half the fun of synths, though, is all the analog controls. I haven’t played with any iPhone or SW synths, but I loved the MiniMoog at Radio Shack back in the day. The 2014 documentary “I Dream of Wires” was both nostalgic and eye opening (to the fact that dev work is still being done!). One button color pattern making would be something to try.

My only experience with multicolor was the functional but so too-simple 4sevens Quark RGB (no color mixing, hideous 18650 tube, but the RGBW MC-E was so cool), and the Technology Associates Rav’n 5mm RGBY 3-button 3-AA box with a big-ol’ DIP uC in it. Neither did mixing or had more than one brightness for each color, but they were cool. There was also a button cell keychain Rav’n which qualifies as 3-color 1-button and I think mixed. If I was going to do the hardware design, though, I’d have to use 4 SMPS current regulated supplies. PWM is, to me, digital software intruding into my analog hardware. Not that the “S” in SMPS isn’t sort of digital, but once you’re in the MHz range, without RF circuit layout, digital starts becoming analog. Anyway, when I’m out of work, I don’t have the budget or inspiration to develop my own flashlight (or a workbench), and now that I am working, I don’t have the energy to do more than some battery and flashlight testing for the fun of it. So I’ll just enjoy what’s out there and affordable. Affordable is relative to income, of course, but just out of principle, I can’t ever see spending over $100 on a light. I think I did for a Surefire Aviator A2 back when THAT was the height of technology, and it is still a wonderful shelf-museum piece. Not sure what the “budget” in BLF means. From this thread it seems like $200 for a light would be unthinkable, but four $40-50 lights is, well, it’s only $40 at a time, right?

As far as kids and lights, depends on the kid. I could have handled a D4 and LiIon charging at 8 or 10. A good portion of adults, most evidently those who like to vape but don’t care to learn anything about electronics, shouldn’t be trusted. Then again, I did melt a carpet with a 7W Christmas lightbulb. As for eye damage, just going by casual knowledge, a flashlight doesn’t begin to compare with even a low-ish power laser. The D4 certainly isn’t going to burn your retinas in the fraction of a second you can look at it without wincing and closing your eyes, only to find it still is uncomfortable with them closed. If you stick it a few inches from your eyeball for several seconds, though, you’ll probably burn your iris, and eyelashes, and eyelid, and it’d really hurt. By the same token, I’d hope kids don’t stick their noses a few inches from a stove burner.

My sweetie does have a 4x7135 convoy C2, but remembering that it has 3 modes is about the limit of what she wants to learn about a flashlight. i.e., You turn it on and it lights up, and there’s something wrong with the light if it’s more complicated. It’s not smarts, it’s attitude. I handle charging or replacing the battery. If I were to give a light to any non-flashlight person, of any age, it would not be 18650. Internal LiPo, maybe. AA or two? Definitely. Sofirn makes a lovely AA light, maybe the SF10. or a Fenix E05. That’d be about it. Hard enough to get someone not used to it to charge a NiMH. Come to think of it, USB charging would be easier to get them to do, assuming they charge something like a phone or camera on a regular basis. USB internal 18650? Eh… maybe.

I always kinda wanted a Techass Rave’N. I still have an ErgoXRay kicking around.

Synths are great fun. I prefer the hybrid variety though, mostly-analog signal path with digital controls and digital effects, so it can get a rich sound but also can load and save presets and automate tonal changes over midi.

But, that said, I also have a Microbrute… which is pure analog to the extent that it uses CV controls, and patches must be dialed in each time they’re used. Sometimes it’s fun to just loop a pattern while painting a rainbow of sound with the knobs. Other times I use it in more carefully-composed songs, usually as a bassline.

Anyway, I’ve also found analog sound synthesis concepts useful for lighting purposes. The candle mode in Anduril is implemented as a 6-oscillator synth where three of the oscillators modulate the other three. And the lightsaber UI works as a 2-oscillator synth moving through a 2D colorspace (hue and brightness), with one wave oscillating between two arbitrary points in the space and the other wave oscillating strictly up and down along the brightness axis. It’s not the fanciest synth design ever, but I wanted to keep it relatively simple since there’s only one button to work with. Despite that, it’s still far more flexible than any other saber engine I’ve encountered, and the 12 profile slots almost feel like they’re not enough.

The interface for that is:

The current pattern is shown real-time during configuration to make it easier to adjust by feel.

Thanks goshdogit. I will hunt for deep carry clip.
Won’t have to be captive or pretty

TK, I cringe to think of the relevancy of “pew pew pew” should that cat backfire…. :wink:

Read this at just the right moment - needed a good laugh

Thanks Dale

Lovely image. Use cats with caution.


Was thinking about efficiency after getting my (2011) ZL SC600w (before it was Mk Anything) and using it for my typical make-a-sandwich and take-a-shower brightness tests. The D4 is OK at it’s IIRC ~125 lm CC mode, where it might get just noticeably warm after an extended run. Around 300lm is really better, though. Not sure where that is on the D4, or how long it’d run, but it gets darn hot (45+C?) if below the thermal control threshold. Double 350mA plus fudge for FET current efficiency loss maybe 1000mA, ~3 hours? The SC600w does 284 for 3h (NCR18650A 2900mAh) and gets “darn hot” too. Maybe they’re not so different. I was surprised the ZL was so hot at such a “low” level. (Turbo is 645) 2011 XM-L vs 2017 XP-G2. Thought I’d compare to the latest maybe most efficient 18650? ZL MkIV+ makes 358 for 5.1h (3400mAh?) and… after a half hour was cool to the touch. (705lm is 2.5, maybe PID kicks in). There’s been a lot of mention of how bright a light can be with this size and ability to dissipate heat. The next level up is 705lm (2.5h) with PID theoretically active, but just tailstanding the whole light gets just a little warm (40C). So at least 800 should be possible if 45 or 50C equilibrium is allowed. Which, explains why I was initially scowling when I set the D4 to, IDK, 400lm or so and it stepped down below 100 in a few minutes (with the default 45C limit, which is really 35C on mine).

The tint is so much nicer than either SC600, though. Only thing that comes close is the Jaxman E2L Nichia triple, which at half power (1050mA) gets a little warm (no idea of output, runtime must be about 3h).

I use Armytek Wizard clips on lots of my lights. It’s my favorite deep carry pocket clip. I grind off the ‘ears’ and sometimes thin the arms to better fit the grooves on the battery tube.

Here’s what it looks like on the Emisar D4 in bezel-up configuration. With some grinding, it could fit even better. Click the images to embiggen.

It’s too long to allow the D4 to tailstand if mounted to the rear groove, but could probably be mounted around the thicker part of the tube.

I ordered a dozen directly from Armytek, but they shipped from China. Other vendors sell them individually. Killzone Flashlights currently has them in stock.

These are a tight fit on most lights, so be aware that the flashlights could be scratched. I don’t mind but thought it was worth mentioning.

We like the same clips.

No such button, but i managed to like” your post somehow hehe
Did the clip worked out the box? I had to bend mine off to just attach it! Too bad angle

I also use the skilhunt h02 clips. I carry bezel down. And I have used epoxy on clips in the past.

Of the options, which led would be the best for flood?

The Nichia 219C option.

That’s a nice looking setup. Is that innertube? I take it you bite that light… :laughing:

Yep, the Nichia will be the floodiest under the stock (10622) TIR. However, I think it can vary with specific LED/TIR combos.
Some emitters can perform differently and be more or less floody depending on the specific TIR being used.
If you look at the Carclo website, it looks like the XP-G2 is actually the more floody option with a 10623 optic (or rather, still a narrower hot-spot than the Nichia, but with more spill).
That said, if you are looking for max flood, I’d go with the floodiest TIR optic (10624), and find out which emitter gives the best performance under it. YMMV.