Quark head disassembly / internals / emitter board - PICTURES!

[uplite]

Inspired by the folks on this cpfmp thread, I finally opened up one of my Quark heads! Figured I'd share my notes and pictures for anyone else who wants to mod their quark.

First up: The bezel is screwed onto the head and stuck with blue loctite. Someone already posted a picture of this:

The bezel was very easy to remove. I wrapped two heavy duty rubber bands around the bezel and around the base of the head, to protect the knurling and finish. Then I used two pairs of pliers to unscrew the bezel from the base. If you have a pair of small strap wrenches, this would be even easier. :thumbsup:

Closeup pictures follow...

Quark head with bezel and reflector removed. A black plastic alignment ring fits around the XP emitter board. The reflector snaps onto four little tabs on that ring, around the LED. The two holes in the alignment ring are useful for picking it up with tweezers. Or it might just fall off when you pull the reflector. It is not glued or screwed on. :thumbsup:

Quark reflector from outside. Nice chunk of milled aluminum, 19 mm diameter, 12 mm deep, 2.75 grams. This would make a good heatsink if it was attached to the emitter mcpcb...but it is thermally isolated by the plastic alignment ring.

Side view of Quark head with bezel, reflector, and alignment ring removed. You can see some of the blue loctite still on the threads.

Top view of Quark emitter board. XP-E Q3-5A emitter on a 12.5mm round metal core circuit board with hand-soldered power connections (red & black wires). I assume the white wire connects the driver circuit to battery ground via the Quark body tube? Haven't removed the board to see.

Closeup of MCPCB edge. The white goo is a rubbery compound that holds the board on the head, but allows it to wiggle about 0.5mm so the alignment ring can do its job. It's a good cheap way to get the LED aligned with the reflector. Unfortunately it sucks for heat management. There is no direct contact between the emitter board and the metal plate underneath.

Closeup of white wire connection to the body tube. Is it soldered to the anodized aluminum, or is the anodize removed underneath the solder?

Blurry photo of Quark head parts: bezel, o-ring, window, reflector, alignment ring, emitter/driver/base. The o-ring is actually the one from the base of the head, not the one inside the bezel. I'm leaving the bezel o-ring in place until I have a GITD ring to replace it, just in case I tear it.

That's all for now. If enough people want to see it, I might unsolder the wires, remove the emitter board, and try to remove the driver. We'll see. This Quark AA gets very little use compared to my Quark 123, but I still like it. I don't want to kill it on the operating table!

Share & enjoy!

-Jeff

 

[Billy Berue]

Excellent, Jeff. I've been wanting to pop mine open and have a look inside, but I haven't had the courage to crank the bezel apart. Thanks for sharing.

 

[NoFair]

Nice pics.

The white "goo" is probably thermal paste to improve heat transfer from the emitter

 

[ajray]

Very cool. Thanks for that.

I'm a EE student, and a friend of mine just introduced me to these (as well as CPF), and what attracted me to this light is the programmability.

I've been wanting to open up the other end of the head and look into the circuitry. I'm curious to see whats going on in there (asic/rom/???) and see if i could possibly read the memory (and/or reprogram it). I wouldn't be surprised if the broke the r/w fuse to make it impossible to read memory, but you never know. I can see the 5 debug headers in the end of head, and I'd like to reprogram the delays. Now that i'm familiar w/ programming it, I'd like shorten the timings. 3 twists in a shorter period (instead of 4) and 2/5 secs to program instead of 3 and 10.

I look forward to seeing that if you do that, otherwise I'll be sure to post my photos here when i get around to that.

 

[lolzertank]

There's a solder connection to ALUMINUM!? I hope it's not a critical connection.

 

[dudu84]

:goodjob: Jeff, :thanks:
I'm curious to see what's under the hood too

 

[DimeRazorback]

Cool!

I wanna do it now, just cause.

 

[Dan FO]

There's a solder connection to ALUMINUM!? I hope it's not a critical connection.

There is probably a screw or rivet under that blob of solder.

 

[Crenshaw]

Okay you have two bits slightly misunderstood.

firstly, the Reflector cannot contact the emittor, because it is a conductor, which means that is will short out everything if it isnt insulated from the rest of the light.

secondly, the white gooey stuff is most likely thermal paste, maybe not artic alumina, but something similar no doubt, which aids heat conduction from the LED board to the rest of the light.....:tinfoil:

and my guess as to WHAT that white soldered connection is, is that it is to conduct heat away from the circuit board. anyone who has ever modified thier lights knows that the circuit board can heat up as much as the LED.... ground wires are generally green and yellow right? i may be wrong, but thats my guess..

Crenshaw

 

[Nake]

That white wire connection is like the Fenixes, for turbo mode when the tube contacts the circuit board. Even the Fenix wire is white.

 

[Crenshaw]

That white wire connection is like the Fenixes, for turbo mode when the tube contacts the circuit board. Even the Fenix wire is white.

Oooo....

you learn something new everyday

although, shouldnt it be a the other connection, rather than the turbo? seeing as if you tighten the fenix head, it goes into turbo, meaning that the body has to contact the pcb -ve contact, and when its loose, the body only connects to the head's threads, which i assume is completed by that white wire.


Crenshaw

 

[EngrPaul]

This design looks all looks good to me! (although I prefer to see a cleaner PCB).

Like someone said, there's probably a solderable screw or rivet under the white lead solder joint.

You don't want the reflector getting really hot, this could lead to the silvered surface separating from the aluminum or hairline cracking near the base. It does get warm even without direct contact.

You definitely want the heat leaving the emitter and going down through the mid-body. The "white goo" under the board is perfect for that, as long as there is pressure keeping it down against the mid-body.

It's good to know that the lens is not terrible to access. The lens is very close to the front of the light compared to others, I worry that it's subject to scratching and breakage.

The excess solder from under the emitter looks like it could interfere with the plastic centering fingers. Hard to tell whether there's relief on the underside for those solder blobs.

Solid, simple design. I couldn't ask for it to be better. :thumbsup:

 

[Dan FO]

It's good to know that the lens is not terrible to access. The lens is very close to the front of the light compared to others, I worry that it's subject to scratching and breakage.

It's sapphire on the outside of the lens so that should keep scratching to a minimum.

 

[Pekka]

Great autopsy pictures. But are there really only 2 O-rings? IIRC there was supposted to be 5 and I can only find 2 on mine outside the head... switch's rubber boot doesn't count as one, does it? :thinking:

 

[Nake]

Oooo....

you learn something new everyday

although, shouldnt it be a the other connection, rather than the turbo? seeing as if you tighten the fenix head, it goes into turbo, meaning that the body has to contact the pcb -ve contact, and when its loose, the body only connects to the head's threads, which i assume is completed by that white wire.


Crenshaw

That's correct, I had it backward. Oops.

 

[uplite]

A few followup comments... :wave:

The white "goo" is probably thermal paste...

Right. And it clearly works well enough in this light. :thumbsup: It just seemed too thick and rubbery imo, preventing the mcpcb from making optimum thermal contact with the base. BUT I just scraped some from the edge of the board, and it doesn't feel so rubbery. Perhaps something else is underneath the pcb...or perhaps the wires lend that "mounted on rubber" feeling to the pcb. :thinking:

the Reflector cannot contact the emittor, because it is a conductor

A metal reflector could touch the top of the mcpcb. You just have to avoid the solder blobs and traces. However...the Quark design works great as it is! :thumbsup: These were just my first thoughts when I picked up the reflector..."wow, that's a good heatsink!"...followed by..."oh wait, no, it isn't on the thermal path".

You don't want the reflector getting really hot, this could lead to the silvered surface separating from the aluminum or hairline cracking near the base.

Good point. My guess is that the light would be too hot to hold long before the reflector breaks down, but I dunno. I haven't touched the shiny surface. Don't want to scratch it.

It's good to know that the lens is not terrible to access.

Yes, this is great! My main reason to open the head (besides idle curiousity) is to put a glow-in-the-dark o-ring between the bezel and the window. The window is 19mm diameter. I just ordered some 20mm GITD o-rings that I hope will fit. :candle:

The excess solder from under the emitter looks like it could interfere with the plastic centering fingers. Hard to tell whether there's relief on the underside for those solder blobs.

There is no molded relief, but the solder beads have created their own mirror-image impression in the underside of the centering ring. The fingers are not pushed up enough to matter.

are there really only 2 O-rings? IIRC there was supposted to be 5 and I can only find 2 on mine outside the head... switch's rubber boot doesn't count as one, does it? :thinking:

Yes, 2 o-rings in the head, and yes, I'd count the rubber boot as the "fifth o-ring" for waterproofing. BUT there really is a fifth o-ring in the tail, inside the retaining ring, around the spring:

It's sapphire on the outside of the lens so that should keep scratching to a minimum.

When I hold the lens to my eye at an angle, it reflects lights equally from either side. I can't tell if it has the same coating all over, or no coating at all. Anyway, it's nice to know that we can replace it easily. :thumbsup:

I'm a EE student...I've been wanting to open up the other end of the head and look into the circuitry. I'm curious to see whats going on in there (asic/rom/???) and see if i could possibly read the memory (and/or reprogram it).

Go for it! :twothumbs I'm sure you have better tools and soldering techniques than me. Here's a possible approach:

1. Unsolder the 3 wires and remove the emitter mcpcb. This is easy. Worst case, you might need some aluminum soldering flux to resolder later.
2. Look into the two holes in the head baseplate with a bright light. Does your school have a fiberscope? That would give you a good picture of what is inside before you try to remove it.
3. Try to press the driver board out of the bottom of the head using two probes stuck in the holes on top. This is the tricky part. The driver is probably glued in. If it is, you would have to back off before you break it, and apply heat and/or a solvent to break the glue.
4. If the driver is potted, cut away the potting material.

FYI, the Quark brochure has a cutaway view of the head that shows the driver. Here is a zoomed and brightened crop of that picture:

:thumbsup:
-Jeff

 

[Frank_Zuccarini]

Thank-you, Uplite, for doing this for us.

Frank

 

[moses]

Just found this thread. Not sure if you are still following but I have a question. I'd like to SWAP the emitter board between two quarks. Is that possible? It seem to be soldered solid to the body?

Thanks,
M

 

[Nake]

It's held to the heatsink by that white thermal goop. I can move the emitter board back and forth a little.

 

[uplite]

I'd like to SWAP the emitter board between two quarks. Is that possible? It seem to be soldered solid to the body?

I think...you can just de-solder the red & black wires, pull the board, and solder the new board in place. Use a high-wattage soldering iron. That metal core board will suck heat from the solder joints PDQ.

If you want to improve the design...you could drill & tap two M2 holes in the head/heatsink, and screw the emitter board down tightly for an optimal thermal path.

After you remove the emitter board, you might try pressing the driver out of the head via the two wire holes. Heat the head in a skillet to 150°C to break the adhesive. Press the board with two blunt probes while still hot. You would probably be the first CPF'er to fully disassemble a Quark head. :twothumbs

-Jeff