First Look: D7 Host (1x18650 EDC) from China Quality Goods (cnqualitygoods.com)

Flashlight Modding and DIY Parts
1

This is a first look on the D7 Host (1x18650 EDC) from China Quality Goods (cnqualitygoods.com)

I paid $14.52 ($12.50+2.02) with shipping included to Germany.

This host is also used in complete flashlights like in the:

OEM D8 CREE XM-L U2 CW/ XP-GBWT R2 7A NW 1x18650 5-mode Stright Tube Flashlight from CNQG ($24+$2.02=$26.02)

and the

EDC 18650 XP-G2 R5 / XM-L U2 850/450lm Flashlight from IO ($26.00-28.99 depending on the options)

So you already have the choice of pre-built versions of this light/host (with a starting price of ~$26 shipped):

Cool White - XP-G2 R5 2B 3/5mode Nanjg 101-AK-A14 (4*7135) driven at 1.4 amps

Cool White - U2 1A ( +$1.99 ) 3/5mode Nanjg 105C (7*7135) 2.45 A

Neutral White - T6 3C ( +$2.99 ) 3/5mode Nanjg 105C (7*7135) 2.45 A

Cool White CREE XM-L U2 5-mode (Hi>>Mi>>Lo>>Strobe>>SOS) 7135*4 (also Nanjg 101-AK-A1 ?) Current:Max output 1.5A

XP-G BWT R2 7A LED "NW" 5-mode (Hi>>Mi>>Lo>>Strobe>>SOS) 7135*4 (also Nanjg 101-AK-A1 ?) Current:Max output 1.5A

So building an XP-G(2)/XM-L light out of this host isn't a very rewarding undertaking, since you probably end up with the same/higher total cost as the pre-built versions. So my idea is to build a 1X18650 EDC with a Nichia 219 High CRI emitter and a Shiningbeam 1.4 A 3-mode driver.

Here some detail pics:


Traditional white box

I like the grey anodizing a lot, since it is a nice change from all the black lights I own. It's a little "lighter" grey then my XinTD SC-82 (C8 Style)


Stainless steel bezel without crenelation... so it's "factory-foyed"...perfect

All parts came bubble wrapped to prevent transport damage



The small white distance holder looks like it was originally designed for XR-E emitters... I have to see how well that works for the Nichias

Very light orange peel reflector ... so I expect the light to be somewhat on the floody side.







Screw-in brass pill with screw-in driver retention ring... never seen that before on any of my lights.

(Match has also seen that on other hosts from CNQG)

Front of the host with threads for the pill and the bezel

Pre-assembled tailcap/switch assembly and anodized tailcap threads....

BTW this will prevent the use of some Solarforce clips (at least without some modification... see comments below)



Glow-in-the-dark tailcap that perfectly tailstands, even when the lanyard is attached. The grooves make a little easier to activate the switch.



Size comparison with the Trustfire TR-801, Shiningbeam S-Mini and the Ultrafire UF-2100

I would put the overall machining of the D7 closer to the S-Mini than to the TR-801 or the UF2100

Comparison to the Ultrafire C8, Solarforce L2T (P60) and the Klarus P1A (1AA)

I will reserve my final judgement until the host is filled with "life", but so far it's looking good.

Do you like this host? Give it a star rating:

2

Lookeen good mang

3

Doesn't prevent but makes it little bit harder... small spacer ring inside tailcap and it will work.

Mine is like this, ugly but working.

4

That seems impressive Vectrex thank you

I think that would be a good start for a noob like me to try have a custom assembly together :slight_smile:

5

Nice ghetto mod... hehe. Sure, you can you "raise" the switch retention ring in one way or the other.... maybe even by putting a paperclip-wire-ring or an O-ring between rubber tailcap and the housing?... This way you would have a "nicer" contact surface on the inside.

6

Nice. What's the reflector depth?

7

reflector is 11.5-12 mm in depth and 20.5 mm in diameter

This how the Solarforce clip would look like attached to the host, but it interferes with the O-ring position to some degree.







8

Host Seems Excellent and I Always Wanted a High CRI Light

Well, I would agree that if all you want is another XML light with this form factor, it probably wouldn’t pay to get involved with this DIY project. On the other hand, I have been studying the difference in color rendition accuracy between the Nichia and the XML emitters – even the so-called “Neutral White” variety and the difference is stunning. If you look at the curves for the XML, there are two pronounced spikes and a huge dip in the curves. The curve for the Nichia High CRI emitter is much smoother – in fact, dramatically so. I would love to have an EDC light like this that actually allows you to see things as they would look in natural light. As far as I know, there is nothing like that on the market.

About the host: To start, I love the color and the overall appearance. The parts seem very well made and it has several other desirable features. I agree with Foy about Bezel design and believe that the SS smooth design is the best. The tail standing capability and switch access seem to be optimum as well. Even the lanyard mount won’t get in the way. The anodized threads are a nice touch. I wouldn’t expect to see square threads on a light at this price but the conventional threads look clean and well machined.

They’ve also added some welcome additions to the interior. The screw-in reflector should help with heat dissipation and eliminate the problems with centering that a lot of lights have. The screw-in pill and the screw-in driver retention ring are both welcome details. The alternative P-60 press-fit is getting a little old, particularly with the trial and error aluminum foil requirements. With all the threaded connections (reflector, pill, driver retention ring) I would bet this light has surprisingly good heat dissipation for its size and weight.

QUESTION #1: I understand that the Nichia LED would not be as bright as an XML. I have to admit that I usually run my lights on medium but what do you expect for max output? Personally, I would like to see at least 300 lumens.

QUESTION #2: Regarding the SolarForce clip, the photo of the assembled light with clip attached looks great. It’s only when you remove the tail cap that you see how close the clip is to the O-ring. I would guess that this would not present a meaningful problem. What do you think?

QUESTION #3: Good luck with this project. When do you think you’ll have a completed light?

FINAL COMMENT: Thank you for sharing this. My interest is peaked and I’m anxious to see how the light turns out.

Regards,

Bob

9

So Vec, you’ve finally entered the exploding world of lithium cells?
:slight_smile:

Good on you!

10

Well Don... you have been away some time, but not this long. ;-)

You might even remember posts like this.

You are right about that I'm still a big Eneloop AA NiMH fan and only surpassed in my resistance to integrate Li-Ion cells in my daily use by Old-Lumens. ;-)

I just looked up the purchase date for my batteries and I purchased 2x 2400 mAh 18650 HiMax batteries on 11/1/2010 and 2 x XTAR 18700 2600mAh batteries on 05/31/11 and these are still my only LiIon rechargeables.

@Kumabear: Thanks for taking interest. The main difference is the wavelength distribution over the whole visible spectrum and it seems to be more even on the Nichia emitters than opposed to the neutral white Cree or even the high CRI offerings from Cree. (see page 10 of this document and page 3 of this document) It's always hard to explain why a Nichias 4500K emitter have a better color rendering than a Cree 4500K emitter, because often people think they have to have the same attributes. I think to some degree Cree sacrificed the CRI to have better efficiency/the highest output available, because many people in decision making positions will favor the concept of pure, raw output power much more, than the concept of sacrificing some lumens to get a superior color rendering index. I have written a little bit on my take on high CRI lights here and why I think that even the industry standards on measuring the index are outdated. (it's not the "real" or "subjective" CRI that's measured) Probably the most convincing are beamshots like these.

I expect the output/efficiency to be in the realms of NW XP-G emitters... maybe ever so slightly below that. The Nichia has a B10 Flux Bin(100-110lm@350mA) so I would put it the range of an CREE XP-G Q4-Q5-R2.

I hope to be able to finish the built in the course of this week, but life often gets in the way of doing things like this... so no promises.

Regarding the clip: The O-ring and the potential lack of being watertight is not the only problem, but also that the anodized threads will prevent the light from closing the electrical circuit. (see workarounds in posts above) But I agree that the light looks killer with the clip attached and easily would make it my best looking light in my 30+ light arsenal.

11

Yes, there are some similarities to the Xenos, but I think the D7 is actually a better design. But it is probably too early to make a call like this now.

@MarioSS: If the driver retention ring works as intended and you have a pre-wired driver like the one I posted above, you could get away with only 2 solder dots to the led star. That would make it one of the most modding-beginner-friendly-hosts ever. The only downside might be, that I have seen Nichias only on 10mm stars/PCBs so far, so the +/- contacts are pretty close to the LED dome.

12

Yup, and 20mm is too large:

https://illuminationsupply.com/nichia-nvsl219ath1-4500k-92-cri-b10-bin-20mm-star-p-159.html

I wonder why there are none on 16mm available.

13

They do sell blank 16mm stars, why not just reflow it?

14

The s-mini clip works on the 2100, so I’d guess it should work on the D7/D8. The 2100 scratches fairly easily so I would be interested to know how the HA3 holds up to it if you try it. I’d really prefer something deep carry similar to the 4Sevens clip Match seems to like.

15

I can do it, you can do it, but might be a little difficult for beginners. :P

16

Yeah, reflowing is not the procedure you want to start with, if you have never soldered before. And the 10mm star should actually be OK, if the heat transfer between star and brass pill is OK. If you have doubts about the heat path, you can always pot the driver compartment to give it additional thermal mass. I have been running XP-Gs on 8 mm stars for some time now, works out great.

That would not be an option for me, because that clip is not sold separately (or is it? update: oh it is.... nice ;-)) and I am not leaving the S-mini "clipless" since it's my most used light recently.

17

Glad you found it . . hope you’ll at least try it and let us know. From the pics I see the D7 is shorter than the 2100. How do the diameters compare? Looks closer than I expected. Oh, how do the tube walls compare, too?

That’s where I got mine and it seems to work well enough . . I just prefer deeper carry when the pocket is more toward the front than on the side.

18

Yeah, the expensive shipping cost to Europe would make it a ~$9 clip for me and so I would wait for my next SB purchase to order it, if it's still available then. Also I think I wouldn't like the black/grey contrast very much, especially since the SB clip is so massive. I will probably try one of my fenix clips at some point. Regarding the potential scratching: I found a nice trick for putting on the clips, take a laser printer overhead foil and wrap it around the flashlight. Put the clip in place and pull the foil out. Those foils are thin and very rugged. I was able to to equip a Xeno light with a Fenix clip this way without scratching it. It seems that Kreisler banged up the HAIII on his Xeno without using this trick. I haven't found a similar trick for taking off clips, so the S-mini clip is staying where it is for now.

The S-mini is 22.5mm at the knurling and 22mm on the flat surfaces and the D7 is 24mm in diameter. The inner diameter of the tube is 18.9 mm, so the walls are roughly 2.5mm in thickness at the battery tube. The D7 is therefore better suited for drivers beyond 1A than the S-Mini.

19

Can this host take protected 18650? Roche F12 will not take most of the protected batteries. I’m kinda concerned with the battery fit of these slender tube flashlight.

Edit: just noticed the post above me state 18.9mm inner diameter. Cool!

20

looks like the perfect host to try out some 20mm TIR optics :slight_smile: Looking forward to the finished product.