#3 Zoomie mods

18 posts / 0 new
Last post
bikedude
bikedude's picture
Offline
Last seen: 4 months 2 weeks ago
Joined: 08/08/2014 - 07:35
Posts: 229
#3 Zoomie mods


Driver: Potted with a 50:50 mix of ‘Coca’ brand thermal glue and graphite. Doesn’t really harden, this glue is a white substance very similar in looks to other silicon-based glues and pastes. Dilution doesn’t seem to be the problem.
This driver probably doesn’t need potting, but I needed to test the glue, and graphite as a substitute for silicon carbide. I intended to switch the stock driver with one of higher output but didn’t find anything for this particular diameter (see other thread http://budgetlightforum.com/node/35281) that would work with NiMH AA cells.
Pill / heat sink: grabbed it with slim round nose pliers and twisted counterclockwise to extract. Rested the emitter (reflector had come loose anyways after I dropped the light on a carpet unintentionally) on two toothpicks to secure it while de-soldering driver connector wires (red/blue for plus/minus pole respectively).

A 10 Euro-Cent piece (see http://en.wikipedia.org/wiki/Nordic_gold), ground/polished using files/WD40 or similar oil for better grip, grit 2000 whetstone, then sawn and filed to correct size of slightly under 13 mm was used for a heat sink. Doing such pieces without a lathe is a pain in the ass. Two openings were filed for the wires, then de-burred using grit 400 sanding paper. I didn’t achieve a tight fit so I glued the piece in place using 5 minute epoxy mixed with some graphite, using the old LED on top to keep it perfectly plane while aligning it with a stick.

Used cotton thread to get the wires through.

Driver reinsertion. Slightly uneven, the HS was almost too thick.

LED: Phew. Not pretty, but it works. Warmer tint at what looks like higher output compared to stock. Still missing the reflector at this point, seems to do well enough without it. LED had to be filed down by ~1 mm, carefully to prevent shorting connector pads to case (guess that would have been a problem else, at least in case of plus pole. Don’t understand flashlight electronics all that well.). Held the pill in a vice, emitter in a helping hand made from clothespin (wood, filed to a point) epoxied to flexible tube from dead lighter (similar to these http://www.geekalerts.com/u/Zippo-Flexible-Neck-Lighter.jpg). Soldered one wire to it’s respective pad, applied thin film of thermal glue (‘Stars-922’ this time, actually hardens to thick rubber texture) to emitter base using a cotton bud, the fitted emitter onto HS, twisting to get wires into place. Tight fit, secured it (emitter) with small amount of super glue. Should have taken more photos of these steps.

Glues used, size comparison.

Components:
‘#3’ Zoomie http://www.banggood.com/CREE-XPE-Q5-600-Lumen-7W-Zoomable-LED-Flashlight-1×14500-p-907006.html?p=CV2621791095201409D6 $3.30 using BLF code (sometimes cheaper on ebay: http://www.ebay.com/itm/CREE-LED-Flashlight-3-modes-500LM-Adjustable-Focus-Zoomable-Torch-Light-14500-AA-/161354035813?pt=AU_Lighting_Fans&var=&hash=item2591733265). Or wait until there’s another $2.99 deal for them. Aliexpress doesn’t have them cheaper, although I guess you could try finding one that looks a little different, maybe you’ll get one with a solid pill… probably not.
COCA thermal glue (not a real glue) https://www.fasttech.com/p/1833300 $3.14 / 62 g
Stars-922 heatsink plaster https://www.fasttech.com/p/1341200 $1.22 / 5.2 g
XPE 4-5000K 16 mm value pack https://www.fasttech.com/products/1609/10010167/1883800 $4.58 for 5, $0.92 per emitter – considering the effort involved, a XPG2 or XML might be a better choice.

Beam shots at 1/10 sec, ~1m distance (sloppy, no tripod) .


Emitter lines are clearly visible (to the eye, not on this pic tho), despite the current lack of a reflector. Slight artifacts from the wires, may superglue them out of the way later.
Output, as already mentioned, appears to be slightly higher, presumably due to the better heat management and more importantly the most likely higher bin of the new emitter.

Image mirrors:
http://postimg.org/gallery/1sy0×0p6/33f7b9a9/

Edited by: bikedude on 01/29/2015 - 09:36
hank
hank's picture
Offline
Last seen: 6 hours 16 min ago
Joined: 09/04/2011 - 21:52
Posts: 8767
Location: Berkeley, California

I got one with a solid pill; think it was the Rustu version (I ordered another Rustu and am waiting to see if I remembered correctly)

For the ones with holes in the pills, I bought some copper disks on Etsy, and notched them, but didn’t file them to fit them down inside the hole in the pill,
I let them rest on the rim — so the emitter got raised that much closer to the lens.
That makes the flood a little wider, it also kept me from getting a clear image of the emitter at the far spot setting but that was no loss, nice tight spot just fuzzy-round.

Ledsmoke
Ledsmoke's picture
Offline
Last seen: 14 hours 56 min ago
Joined: 08/08/2011 - 16:05
Posts: 1995
Location: Denmark

Well done. 

But a word of warning though. Graphite and aluminium are not good friends. Graphite has semi metallic properties and as such can and will create what is known as galvanic corrosion (bi-metallic corrosion is also used) where the graphite acts as the cathode and the aluminum corrodes. This is an issue when there is moisture, specifically if chloride is present. And it will be in small quantities from your body heat and sweat. 

But all I am saying is that this might be an issue. For all I know the Coca may encase the graphite to a level where it is a non-issue. I sure hope it does. Because like I started by saying.. It is well done.

Thank you for sharing and I really hate raining on somebody's parade like this, so I hope nothing happens or that it happens so slowly as to be negligible.

~ Ledsmoke ~

Dutch humor:

[quote=djozz]

 I do not think that the BLF-community ben

bikedude
bikedude's picture
Offline
Last seen: 4 months 2 weeks ago
Joined: 08/08/2014 - 07:35
Posts: 229

Hm, that is something to consider. In epoxy it’s fully encased or so it seems, if I rub my finger over it there’s no discolouration. In ‘Coca’ it remains exposed to some extent.

Fritz t. Cat
Fritz t. Cat's picture
Offline
Last seen: 1 year 4 months ago
Joined: 05/07/2013 - 00:33
Posts: 2535
Location: Si Valley

bikedude wrote:
maybe you’ll get one with a solid pill…

The one I was testing today from Wallbuys (chosen as my most expendable light) has a solid pill.

Flashlight designers should look at lighthouses and pottery.
这些谁设计的手电筒应该看灯塔,以及在陶器。

hank
hank's picture
Offline
Last seen: 6 hours 16 min ago
Joined: 09/04/2011 - 21:52
Posts: 8767
Location: Berkeley, California

solid pill— oboy! link?

Of course, reminding myself, the nature of a lottery is, it’s a lottery, wherever … but that’s hopeful

Fritz t. Cat
Fritz t. Cat's picture
Offline
Last seen: 1 year 4 months ago
Joined: 05/07/2013 - 00:33
Posts: 2535
Location: Si Valley

http://wallbuys.com/Product/Mini-CREE-XPE-20LM-3-Mode-Cool-White-LED-Flashlight-with-Clip-Black-1xAA—9836
That was on Deal Alerts and this thread: http://budgetlightforum.com/node/25402

Flashlight designers should look at lighthouses and pottery.
这些谁设计的手电筒应该看灯塔,以及在陶器。

hank
hank's picture
Offline
Last seen: 6 hours 16 min ago
Joined: 09/04/2011 - 21:52
Posts: 8767
Location: Berkeley, California

I’d asked when that came up whether it had a solid pill, but nobody had answered that over there.
Much appreciated.

DBSAR
DBSAR's picture
Offline
Last seen: 13 hours 18 min ago
Joined: 02/11/2013 - 23:28
Posts: 6263
Location: Ontario, Canada

I am curious, the graphite is not causing a short on the driver or any parasitic drain ?
Graphite is a high conductor of electricity, and is used to make “resistor compounds” when mixed with a non-conductive medium.
Also as mentioned above there is a corrosive element to graphite when in contact with aluminum parts too.

That Canadian flashlight guy & Lantern Guru -Den / DBSARlight

bikedude
bikedude's picture
Offline
Last seen: 4 months 2 weeks ago
Joined: 08/08/2014 - 07:35
Posts: 229

Yes, graphite conducts electricity well, hence pencil line mods are possible. It also conducts heat at 150-500 W/mK depending on layer thickness (thinner=more conductive), so it could outperform silicon carbide (=> http://budgetlightforum.com/node/33645), assuming particles dispersed in glue have similar attributes to a thin plane.
Silicon carbide is also a good electric conductor if I’m not mistaken; there is probably no direct line of particles (due to dispersion i.e. glue/epoxy gets in the way) in both cases for current to flow through. I guess I (or someone else) should run a conductivity test some time. As I said, the light is quite bright, brighter than stock despite the warmer tone, so I don’t think parasitic current is present or significantly high.
Potential corrosion is a bigger issue. Wonder how I could test it; put potting mix on some alu, put it in a moist environment, then remove mix after a month or two to inspect alu surface?
If the light stops working due to alu components on the driver (pins?) getting corroded I’ll tell you.

hank
hank's picture
Offline
Last seen: 6 hours 16 min ago
Joined: 09/04/2011 - 21:52
Posts: 8767
Location: Berkeley, California

aside — I’ve got several of the Rustu brand #3 zoomies coming from DX, ordered weeks ago.
At last report they did have solid pills.
My last throw of the lottery dice for the season.

Ledsmoke
Ledsmoke's picture
Offline
Last seen: 14 hours 56 min ago
Joined: 08/08/2011 - 16:05
Posts: 1995
Location: Denmark

@ bikedude: Silicon Carbide has been tested by people on this forum with insulation testers with up to 5000V and shows infinite resistance => negligible electrical conductance in real life applications Smile

~ Ledsmoke ~

Dutch humor:

[quote=djozz]

 I do not think that the BLF-community ben

bikedude
bikedude's picture
Offline
Last seen: 4 months 2 weeks ago
Joined: 08/08/2014 - 07:35
Posts: 229

Ran a simple resistance test via multimeter (Benning MM P3).
No conductivity on graphite/epoxy blob (tested at multiple sites, 2-5 mm apart)
300k-1.3MΩ at 2-10 mm on a film of graphite/coca paste. Tried to get a blob, didn’t use enough, smeared it on cardboard instead.
No conductivity on pure coca paste.

Might be enough to cause issues on other drivers.

bikedude
bikedude's picture
Offline
Last seen: 4 months 2 weeks ago
Joined: 08/08/2014 - 07:35
Posts: 229

So, I ran another test, using wires inserted into the epoxy (and some other glues). http://budgetlightforum.com/node/36617
Epoxy remains viable, everything else is dubious at best.

Fritz t. Cat
Fritz t. Cat's picture
Offline
Last seen: 1 year 4 months ago
Joined: 05/07/2013 - 00:33
Posts: 2535
Location: Si Valley

I can mention here a mod. I made to a #3 zoomie that completely fixed its main problem: its appearance. Lol.
I found a nice looking light from http://www.buyincoins.com/item/1391.html#.VJxSk_9rsXA . for $2.89. I liked it except that it had low output and a plastic pill. So I filed a mm. off the pill of a #3 and shoved it in with heat sink compound. It isn’t a zoomy, but some people are happy without zoom. It is a 1 x / 2 x AA light with an extension tube, so the driver is right for it. It puts out almost as much light with 1 X AA as it does with 2 X AA, with two or three times the current.
Length adjustments took some fiddling. Like other plastic pill lights I have, the threads connecting the head to the body are anodized, as well as the inside of the head. The threaded end of the battery tube contacts the driver, in this case without a spring to take up slack. The metal pill and shortened plastic reflector are shorter than the original ones, so I had to shorten the back of the head and add half a mm. to the back rim of the driver to make secure contact.

Flashlight designers should look at lighthouses and pottery.
这些谁设计的手电筒应该看灯塔,以及在陶器。

WarHawk-AVG
WarHawk-AVG's picture
Offline
Last seen: 5 months 1 day ago
Joined: 01/04/2014 - 06:47
Posts: 5071
Location: H-Town

Question: Will that non-hardening potting compound ooze and leak if it gets warm?

Either way, excellent mod!!!

Fritz t. Cat
Fritz t. Cat's picture
Offline
Last seen: 1 year 4 months ago
Joined: 05/07/2013 - 00:33
Posts: 2535
Location: Si Valley

I haven’t had any trouble with the compound moving. There are specifications on the jar about stability at high temperature as well as about thermal conductivity. This seems similar to a typical application such as between a cpu and its heat sink. I have “silicone” and “no-silicone”. The “non-silicone” has slightly higher specs, so I use it except on non-silicone o-rings.
Thanks.

Flashlight designers should look at lighthouses and pottery.
这些谁设计的手电筒应该看灯塔,以及在陶器。

hank
hank's picture
Offline
Last seen: 6 hours 16 min ago
Joined: 09/04/2011 - 21:52
Posts: 8767
Location: Berkeley, California

If you get the “Buyincoins 1 x / 2 x AA light” check and see if its tube threads are compatible with the #3 zoomie or the SK68.
Some (not all) of these “3W Police” 1/2AA lights can be taken apart, the 2AA tube parts reversed, and they thread into the tailcap end of the zoomie flashlight, and use the zoomie tailcap/switch on the other end —- giving you a 3xAA zoomie.

It’s a gamble whether the threads are actually compatible.