Suggestions for venting the case on my achesalot/troutie inspired headlamp
I'm diving in for my second home-brewed light. This is a mash-up of an achesalot and troutie-styled light; both of which I find very interesting. My design goals for this wearable headlamp project are as follows:
-Sub-100g including battery and strap
-200 lumens with 2 to 3 hours capacity
-Max case temp of 50 ° C in still air/light breeze
-A head size that is as small or smaller than Apex Pro
-Preferably no exterior fins
2 x Cree XPG (wired in parallel)
1 x 18650
1 x LFlex driver
1 mm walled, 25 mm square tubing
Which of the following is likely to provide greater cooling capacity:
2mm holes in the case,
Total volume of metal removed 88 cm3
Flat surface area removed (175 mm2)
Side surface area gained 229 mm2
Net surface area gained 53 mm2
or 2mm slots in the case?
Total volume of metal removed 200 cm3
Flat surface area removed (400 mm2)
Side surface area gained 204 mm2
Net surface area lost (196 mm2)
The piece of angle shown inside the case will have a series of heat sink fins mounted between the holes or slots, so that I can get some extra heat dissipation.
Thank you for taking the time to consider this.
a bit more info on why you've chosen a particular option would be helpful. Eg.
why 2 XP-G if you only want 200lms? 1 XP-G can do that easily (and up to 400lms if necessary).
Why the 10mm optics? they have a bad rep for being very floody with not much throw - a single XP-G and Regina would be far better.
Where's the battery going to go? Why only 1 cell? A 2P battery isn't going to be that much more noticeable on your helmet, if this is where the light is destined for.
If you want to use 25mm OD tubing, you can do this easily by simply cutting a U shape from the tubing, stick LED on one end, switch and power cord on the other, with the lflex in the middle. Then slide into a 44mm long piece of tubing, attach with screw through a cateye clip, then use a cateye helmet mount.
That plus a 2P battery should get you in around 120-140g (or 80-100g for a 1P battery) and 350lm/2h. It'll be a lot simpler to make too
matthemuppet - if he's going to use an Lflex then wouldn't it make even more sense to use an xm-l and drive it at low amps? You would have the ability to drive it higher if needed and the Lflex/xm-l will be designed to work at 3.7v.
The reason I ask is that is the light I want to build using your design along w/ troutie's suggestions w/ his "1". How would you connect the sled w/ a screw through a cateye clip? If the Lflex is sitting on the middle section of the sled, wouldn't that make the screw go through the same section of aluminum that the Lflex is on?
Also, how would you add more surface cooling to run that puppy at 2.8 amps?
thanks in advance.
Thanks for the questions, guys.
I want a very floddy light, at 90 CRI, which is what led me to the XPG and 10mm (XML is out due to color and CRI).
I want to stay inside th 25mm x 25mm box size for use as a working/playing headlamp on a strap.
The 1s1p choice is strictly weight/size driven.
The 2 XPG v. 1 XPG is to squeeze just a touch more efficiency.
200 lumens is the average expected output. The 18650 should be good up to 2C draw, and the housing should be good for just a few minutes of "high beams."
Temp regulation is an absolute requirement to avoid a toasty forehead.
Battery will ride on the back of the elastic head strap (back of my head).
So are you looking at using the xp-g Q2 to achieve the 90CRI
As for cooling I would go for fins. That way when you are stopped, natural convection will cool much better as the air travels upwards through the fins. I think there would be too much turbulence in the drilled holes to be efficient.
Makes sense, Brad, thanks. Slots are easier, too.
Yes, the HI CRI Q2s are currently making their way from OZ.
As an aside, my absolute favorite light beam of all time is a Surefire C3 with the P91 incan module, which is what I am trying to duplicate here.
Seeing the beamshots posted here, I do agree with Matt that I would prefer a Regina, but I just don't have the room in this one.
If you're putting heat sinks on the back of the L angle Al and you're only driving two xpg hard enough to get 200lm total then i would completely remove that section of the box and leave the heatsinks exposed so they work the way they are designed.
I prefer your slotted version.
The 10mm Carclos will give you the floody beam you are looking for.
The sub 100 gram mark seems to me a bit tough to hit. I have 2 dual XPG designs that weigh 26.4 grams using Reginas, and 27.1 grams using 10mm Carclos. Neither of those has the driver in the head, it lives in the battery pack. The Carclo version could be lighter if it were not designed to run at 1A (fewer/thinner fins). If your lamp assembly comes in somewhere around that weight and an 18650 is about 50 grams, the remaining 25 grams for head strap, cable, switch, etc. seems tough. Good luck, I am looking forward to seeing this project progress.
ah, gotya, understand now I'm guessing that you need that tint because of what you're working on?
Re Andy13 - the lflex will go like so:
switch II__ __lflex II LED
so that it's on the opposite side of the piece of metal the LED is screwed to (screws cut flush of course). Sending the LED wires through will be a bit of a squeeze, but there should be enough room in the corners, either side of the lflex.
Then the bolt from the Cateye bracket would go through the bottom of the U to hold everything in place. I'd love to make a single XM-L for myself, but there's no way I'd be able to sneak it past my wife when I already have a very nice helmet light and I'm working on a twin XM-L bar light
Matt, I just saw the design you are describing. I like it.
Is it OK to rivet the 10mm star (PCB) to the heat spreader instead of gluing, screwing or taping it down?
I will, of course, put silver 5 thermal compound to put under the PCB before applying the 3 mm aluminum rivets.
I have some arctic epoxy inbound with the XPGs, but now I don't want to commit to these optics 100%. I might be able to squeeze a 1 XPG and a Regina into this box, and use the other LED for ???????
No reason why pop rivets wont work they will pull it down nice and tight
as long as you can rivet to the half holes in the board with out it moving out of alignment
Put the second rivet into the hole before you start to apply pressure / tighten the first rivet. This should help locate the star properly and stop it from being pushed sideways. If you are confident enough I would then also give the first rivet a single squeeze on the gun, slide the gun off and put it on the other rivet and do the same thing. This will snug them up without getting too tight. Once you're happy with the positioning of the star and that it hasn't moved then do them up tight enough that the rivet shaft snaps off as per normal.
Good advise there Stu
and what could be a good method of fixing stuff to these thin tube builds with out drilling and tapping
how waterproof is a rivet when done
A correctly formed rivet will be fully waterproof. Sometimes the rivet shank can pull through, check you have the correct nose on your rivet gun. Also make sure the hole is the correct size. A rivet is designed to expand in the hole, as well as on either side. The amount of rivet sticking out should be the same as the diameter of the rivet. is, a 3.2mm dia rivet should have 3.2mm poking though your work.
We use a mix of solid and 'pop' rivets on the aircraft and they are all water proof.
In addition to Emu26, apply pressure onto the rivet head when you give it a squeeze, this will ensure it's nice and tight in there!
and be careful when the shaft does snap, I always find the rivet gun tends to "bounce" when this happens and you don't want it to bounce onto the dome of your nice new LED
In this case, masking tape is your best friend! Just don't rivet over the tape...
thanks! Though it's a collaborative thought exercise with Ofroad'bent, so I can't take all the credit. I think he's planning on making a single XM-L/ lflex that way, so hopefully we'll see a build report up in the near future.
Originally Posted by P220C
Work Progresses - (Is it an extra gravity day?)
Work progresses on my "thinking inside the box" project. Still trying to maximize surface area at minimal weight gain without resorting to exterior fins.
The box size is 25 mm x 32 mm, and I calculate a total of 10 sq. inches of surface area. I added texture to the fins by running a file on edge at 45 degree angles. Looks like really small diamond plate, if the diamond plate factory workers were dropping acid on plate making day.
It's really hard to win this game, when the strap alone weighs 28 grams.
On the positive side, the strap, and most of the body weighs 46 grams.
I don't know how much weight the 2 x 10mm PCBs w/ XPGs, Carclo optics, and LFlex are going to add, but it seems pretty evident that I am going to have to either concede on my 100g, or 3 hours (i.e. go with a 14500 battery in lieu of 18650). I will probably just make up one battery "pack" of each (one for running with, and one for working with).
Last edited by P220C; 02-13-2011 at 08:23 AM.
2 XPGs on 10mm round MCPCBs, 2 trimmed down Reginas, 100 mm length of 4.3 dia.cable, .75mm lexan lens cover and o-ring seal add up to 8.5g. The 10mm squares will be a bit lighter as they are thinner than the rounds.
Originally Posted by P220C
Still plugging along.
With the front slots and the fins and box beneath, I was up to a total of 10 square inches of exposed surface area. Total size of the box is still 25mm x 35mm
After playing with this setup, I want more... (really though, who doesn't?) At 1.2 Amps (0.6 per emitter), the case got pretty hot after 10 minutes in still air.
I cut 5 slots in the back, and am going to run some 5mm x 1mm copper strips in there as fins. This netted me an additional 5 square inches, for a total of 15 square inches. Hopefully, this lets me run warm at 1.0 amp in still air.
The inside should still be watertight.
One other change: I got rid of the lexan lens, and moved the leds forward, and am going to use the carclo lenses as the main lens. I will use the MTBR patented 'nano glouing' technique.
The reason I chose to do this is that I could tell a difference in the beam's brightness and shape after mounting the leds in the housing behind the lexan. I couldn't open up the hole any, so I just popped the lexan out and moved the leds forward. We'll see how this holds up, but I don't foresee an problems.
The little piece of copper heatsink started the day as a copper pipe clamp. I hammered it flat, cut it with some tin snips, and soldered it up. Turned out surprising well considering I am more accustomed to joining metals at temperatures a few thousand degrees higher than that.
Here is where the little copper piece is going to go. I will shorten the side cover up to leave this exposed to the air.
This leaves me 7 mm to squeeze the in the 3mm thick LFlex.
I could not tell for sure from your pics if the Carclo you are using is the 10417. If it is, you should consider using a cover. The recessed center spot in the face of that optic does a great job of getting muck in it. It is tough to clean out on the trail and a small amount of dirt in there kills what little bit of throw that optic has.
Originally Posted by P220C
You are right about the optics. I didn't think it would be that big of a deal if they got a little dirty.
I found some 1mm thick Lexan at my local welding shop today for $0.89. I'll cut and sand that for the lens.
This is looking to be very similar in design goals to my 2 xpg light which is primarily used for running. The 100g with an 18650 never came for me.
Some pictures here.
Having used it for about a year with a 14500 which came in at 100g or 2 18650 cells 180g I have moved to 3 battery options using lipo cells. They have the advantage of a lot lower internal resistance and stay in regulation better towards the end of the runtime at high drive currents, this is why I was using 2 18650's befor rather than just one. I can also still get full power from the small cell setup if I want which I could not do with the 14500.
350mah all in 85g which is what I carry when the light is a backup.
700mah at 95g which is what I use for training runs during the week, easaliy gives me an hours runtime at the typical 500ma level I use when out training.
2220mah at 150g This is what I have been using for smaller Night Orientearing events this year for biger events I use a 7 xpg setup. This gives me a good hours runtime at 2000ma.
I did not put anything like the effort you are into the cooling, it gets hot in still air but in prictice out in use it never gets more than just warm.
I am impressed by that driver. I've been toying with the idea of putting together an adjustable analog driver based upon a TIP31 for my son's 12v riding toy. I keep thinking how much easier this would be if I just knew how to program a PIC.
I've seen some of your pictures before. In fact I remember seeing the 2 up XPG in parallel wall pictures, which is what led mebto choose these optics in the first place.
Early on in my planning, I saved the image of your 2up on the scale. That's where my 100g goal came from.
I must apologize for not giving you credit, but I couldn't recall where the pictures had come from.
Thanks for the inspiration.
I saw your post in another thread with your light on a scale. Nice job. Haven't seen any pics for awhile. Got any better pics than just the far away top view?
Yeah, been traveling, and the battery and plug were there when I got home. Threw it on the scale after soldering it up. I'll get some better shots later today.
It's able to run @ 1.2 A in still air without getting hot (<50°).
The 2900 mAh battery came in at 45g.
I had planned on velcroing it directly to headband, but am thinking now that a few extra grams in screws for some up/down adjustability would be a worthwhile trade off. Plus the bracket will add some heat sink area.
The LFlex, an 18650, and 2 parallel XPGs is a good match.
Allright, here is the current state of the build. It's been a slow process (lots of thinking, some building, and then some waiting on DX to deliver the connectors).
For the housing itself, I am 10 grams over my intended weight. This was the result of adding the rear heat sink. Considering this extra weight netted me a 50% increase in exposed area, I think it was a good tradeoff.
I'm able to run (with a light breeze) at 1.2 A (600mA per emitter) without tripping the 50º protection.
The spec sheet shows the following:
XPG 90 CRI White, Q2 bin
87.4 lumen @ 350 mA
50º junction - 96%
600 mA - 160%
Lexan lens - 93%
Optic - 90%
x 2 emitters
224 theoretical out-the-front lumens
I'm running the 2,900 mAh (AW's protected Panasonic) at a 0.4C discharge rate, so I should see between 2 and 2.2 hours of runtime at this level.
It's a lot (I mean really a lot) of fun to have the option of kicking it up to 2.5A (408 theoretical out-the-front lumens). Granted, I only get to have 1 to 2 minutes of fun before mean ole' Mr. Thermal Protector turns down the show.
I couldn't be happier with the beam shape for jogging.
One not-so-noticeable difference between the light now and earlier pictures of it is the fact that all of the inner pieces (heat spreader, rear heat sink, and 'L' shaped heat sink) were sanded down to ~1mm to reduce weight.
These JST-styled connectors from DX should handle 3 Amps easily. They also seem to hold together well.
This was a ridiculously tight squeeze. Had I know I was going to be so crammed, I would have used silicone wires instead of PVC (would like the higher insulation melting temperature... just in case).
I was really stumped as to where to mount the switch. I finally settled on carving a notch out of the heat spreader, and using some two-part epoxy to hold it in place. This stuff is rated to 90º so I should be OK. If I had this step to do over again, I would leave some space to allow a rubber boot to fit inside the side cover for water-proofing.
You can't see them here, but I glued four fins (running vertically) between the rear heat sink/spreader and the inside of the tubing. This is in addition to the little homemade heat sinks seen on the side.
Last edited by P220C; 03-09-2011 at 10:51 PM.