Introducing an EDC 2-Cell LED Light (CAD) SPY005

[Data]

Re: Introducing an EDC 2-Cell Light Feeler(CAD) SP

Amorphous

Tons of stuff to think about, thanks. Another vote for a clip, duly noted. Thanks for the kind words too.


Darell

I look forward to seeing you at Ghost Mt. That should be a blast. I am itching to start cutting metal and should have something to show you there.

mounting on a bike: I am a big bike nut, lots of bikes. The first thought that came to me was to make extra tail caps. You take the extra cap and drill through the back of it. It can then be bolted to any off-the-shelf handlebar clamp.

knob: Because my knob is smooth it can not come on in your pocket. Only the friction of one or two fingers will be able to supply the necessary torque to rotate the switch. I have tested the torque of my fingers on a chrome plated 3/8" socket with an OD on .5". It was quite easy to supply the torque needed to turn the rotary switch even if my fingers are wet.

The light will stand on its end-cap, I promise.

I see you are an EV Nut. Cool, I want to build an EV car too. I have two EV motorcycles (48V @ 400 amps) and an EV scooter (36V @ 400 amp).


keithhr

thanks


Cheers
Dave

 

[thesurefire]

Re: Introducing an EDC 2-Cell Light Feeler(CAD) SP

I'll second a few things:

1. It needs a clip
2. It needs to be 'dunkable' at least for an EDC
3. I think 1500 ma is to high, maybe 1000-1100?

 

[McGizmo]

Re: Introducing an EDC 2-Cell Light Feeler(CAD) SP

Dave,
I took another look at your renderings to see if I could make some other heads that would exchange with the stock one and it looks like the rub would be in the heat sink and its thermal bog binding to the head. With your ability to do fancy milling, would it be posible to leave a couple partition tabs that are part of the main body that would serve as mounts for mechanically fastening the heat sink to the body? I think you could effect a better thermal path through the body of the light and having the heat sink rigid might not be a bad idea? You might even be able to pass a screw through the PCB and through the tab and into the heat sink and sandwich the package in a secure fashion.

If you look at the KL4 head, the heat sink is secured by two #4 screws to the head via two small "web bosses" (I have no idea what to call them /ubbthreads/images/graemlins/blush.gif) and the thermal transfer to the head is real impressive!

I assume you are open to at least hearing these off the wall suggetions?!? /ubbthreads/images/graemlins/smile.gif

 

[paulr]

Re: Introducing an EDC 2-Cell Light Feeler(CAD) SP

Wow, incredibly cool, I just saw this thread. My comments:

1) Could you decrease the radius on the back edges a little, so the light can easily stand on its tail without tipping over, even on rough or not-perfectly-flat surfaces.

2) I like the idea of including a charging system, either with contacts or induction. Using RCR123's would mostly solve the back cover attachment problem. Just hold the cover on with screws. Most of the time there'd be no need to unscrew it since you'd recharge the batteries inside the light, so tool-free battery swap is less important. Ideally the charging circuitry should be inside the light so you don't need a smart charger, but can use a simple cell phone-style charger.

3) Rather than an O-ring on the knob, how about some machined scallops.

4) For a light this expensive I think it's imperative that the LED be upgradeable in some reasonable way. It's ok if the upgrade involves soldering and/or a board swap (up to $75, say) but people shouldn't find themselves in a position of wanting the latest LED and having to pay $345 for a whole new light. Also, a non-upgradeable LED means that every LED improvement destroys the secondary market in older units.

5) I wonder if you could add a 3mm red LED to that somewhere. I really prefer red for night vision, over very dim white.

6) There should also be tritium markers, or at least slots for them, maybe in the middle near the tail cap.

I think it will be a bit too rich for me but it still looks like a completely awesome light.

 

[methuselah]

Wow. Beautiful, innovative. Too rich for this cheap s.o.b., though.

How about a serpentine clip? (imagine a squashed "S") One clip, bezel up or bezel down. I've never seen it done before, so It's probably a loser... but it seems to me it could work.

 

[andrewwynn]

great stuff and thanks for sharing with me early-on in the project.. lots of great in put here.. if you make a clip with two holes fore and aft it would take seconds to either remove it or flip it around to appease the uppers or the downers... if you make them close enough it could be possible to have a 4-potition clip and have a forward facing belt held postion (could also work on a hat-band).. this was my idea with the minimighty.. having a clip that can rotate each 90 deg.. well.. now my main clip is only 'head down' design but there is a 1/4-20 hole in the bottom to be able to engineer any type of add'l mount i.e. for bike-mount.. though i'll probably engineer some kind of 'clip' to hold on a bike 'cause the vibration would be pretty bad if held just from the back.

i think that the axial loading of the face o-ring seal would be ok, but my soltion on almost an identical switch arrangment is to use a 3mm o-ring on the inside of the shaft held between the circlip and the body of the light.. this way you can set the compression to whatever you need as you install the circlip and have plenty of slop from a tolerance perspective.. you can still have the outer o-ring just for dust but the 'underwater' tight seal would be only on the 3mm shaft (or 6mm if you hvae a bigger shaft than my design)... it solves a lot of issues especially with tolerances..

another possibility is that the inside of the knob was chamfered and the groove on the face of the light would lock in the o-ring almost as if it were inside the knob.

i think internal charger is a great idea.. i am using a usb minib to get power in.. needs like 4.5mm x 8mm x 9mm internal volume.. and can be sealed on the inside.. since only two pins are used in the jack itself using a gold plated jack means that even if it's dunked in water occasionally it will not be an issue. I will be making an inductive charger for the prototypes of the MM.. not sure if that'll make it to production.. i axed it early on when i realized how much more difficult it is to deal with charging from a DC source.

grayhill makes good switches sounds like you have that part figured out... it would be tough to switch from this design to what i had hoped to use in mine with similar input... i wanted to have a 'clickwheel' ala iPod.. button in the center and wheel on the outside.. my entire control area is about the size of a dime and it's impossible to source a rotary encoder that has a 'click' switch in the center that small but you have the room so i would see if you can find it.. they have them w/o shaft i think so you can use the same design you have now other than the button has to move axially.. the seal could be an o-ring around the shaft .. think of a valve on a water spigot.. you can have axial motion and water tight integrity.. might need to add in a spring.. but being able to leave the light on level 4 and just 'click' on/off would be ideal.. not necessary i like the twist dial a lot.. i use the same on my megasonic and it's incredible and such a 'gimmie' to operate.

I just looked closer at the reflector seal.. you have the seal outside of the reflector that's very nice.. much cleaner view from the side.. what size is the reflector, or is that going to be several options? it's always good to have options like other lights out there where you can screw off one and put on the other.

re: drive levels.. less is more.. there is no need to drive the emitter to excess.. i'm with don on this one.. my design runs the emitter at exaclty factory spec max.. that's pretty high (1500mA for a single K2.. so it'll be good to have the driver prepared to do that.. but with lux3 i'd stick with 1000 or a slight over drive of 1100 or so like i use in the RT4).

batteries.. definitely i would use R123.. if you set it up to be parallel you have 3.7V under load or so and with boost/buck you are golden as long as you can still dim the light (goes in to buck mode when needed).. if you put them in parallel it's very easy to charge with just a couple chips.. it would be sexy to have it be an inductive charge... but you could easily use a couple contacts... the contacts go to the charger not the bat so i don't think there is a worry about shorting.. applying zero volts to the charger input should cause no problem.

keep on keepin on this is a really neat design.. very 'of it's own' and some out of the box thinking... like Harry said.. looks like the year for side-by-side lights.. and each case they are so different from each other i wanna have one of each.

-awr

 

[IsaacHayes]

Re: Introducing an EDC 2-Cell LED Light (CAD) SPY0

Pretty cool. I like the sudden spring up of non tranditional shaped lights. I really like the rotary switch with 6 levels of light. I like the 1.5amp deal too, use it only when you need it. I like set levels of dimming. That way you know what level is best for what situation, and just click to it. Rather than digitaly scrolling up to somewhere and looking to see if it's bright enough yet. Instead you can be like "at this distance I always use 125ma".

Not sure if 2ma is bright enough for use, is it?! It would be nice for the first level to be just enough for reading in pitch dark, like a diffused ArcAAA (arc can be too bright at times)!

 

[Data]

Re: Introducing an EDC 2-Cell Light Feeler(CAD) SP

McGizmo

I am open to hearing any and all suggestions. I would be honored for you or anybody to make replacement heads that are upgrades or offer different solutions for the SPY. So I am keen on your ideas.

With that said, one firm requirement I have for the head that I am making and the body of my light is it not get any bigger on the outside than it is right now. My light, as I supply it, will stay an EDC. The outside diameter of the body where the head attaches is 0.804" and those threads are under 0.750".

Before we can start looking at alternatives I will present you with the full detail of my current design. The current design transfers 80% of the heat directly to the head. The bezel threads are right there too and through them half of that heat transfers to the body. The LED mounting plate does fit into arms that come off the body and so some heat from the LED mounting plate goes directly to the body. I have not yet explained how this part of the light goes together. Nor can you see it clearly in the images I have provided, an oversight I will soon remedy. I will go over this in more detail and capture some extra images later today or tomorrow.


paulr

Regarding your item #4; the short answer is, I agree 100%. The LED will be upgradeable in some reasonable way. The head of my light screws right off, the LED mounting plate comes out and is connected to the PCB with two wires that are about 1 inch long. If you turn the LED mounting plate sideways you can push it right back thought the light body and it and the PCB, with the rotary switch on it, will come out the back end with the wires still attaching them together! So the LED, converter board and switch all come out together as one unit.
/ubbthreads/images/graemlins/party.gif

I am committed to providing an upgrade kit when the K2 becomes available. I have been keeping tabs on its requirements and am trying to make sure my design will lend itself to an easy K2 retrofit. The upgrade will likely involve soldering those afore mentioned wires.

Wayne and I are designing the light to work with rechargeable RCR123s. With the current design you will have to take the batteries out of the light to charge them.

Thanks for the complements.


methuselah

thanks.

 

[Data]

andrewwynn

Please keep working on that clip, I like what you are coming up with!

The reflector will be either a McR-20 or an IMS 20.

I do want to provide options for the head of the light. I do not know what they will be yet. If someone else picks up this task for me then so much the better.
. . .cough Don cough. . .
My SPY005 head will be pretty much as you see it in my pictures.

I second the need to have one of each of the side by side CPF lights made! Bring them on.

Thanks for all your help too Andrew.


IsaacHayes

I have not decided on what the current level of the low setting will be yet. You may be correct that 2mA is too low. I am doing testing for all the powers levels now and will have it complete soon. Thanks for the comments and complements.


Cheers
Dave

 

[andrewwynn]

re: lowest low... the lowest setting i've been able to get stable with a switching driver is about 30mA i think that's almost an established 'given' in LED lighting with luxeons.. pretty sure that's the hard limit on the fatman for similar reasons.. it's a good limit and will give you solid output.. a quick guess on runtime would be 20 to 40 hrs from your light with R123s.. is there any need for more than that?

send me a dimensioned drawing and i can work on a clip designed for your light based on the one i was working on for the minimighty.

-awr

 

[NewBie]

30mA is most definitely not the minimum limit for the output of a switch mode converter.

The limit is more due to the size of the current sense resistor, the gain in the op-amp, and the voltage on the FB node, and the ability to see the actual signal above the noise, which can vary, depending on the skill of the person that did the layout the board. The internal design on the chip also comes into play. Finally, the gain/phase margins come into play, as the right-hand pole is very important to consider, since it changes as the output current changes. As you get into the very small currents, losses in the inductor, and discontious operation can start causing an issue.


Could you explain how you arrived at the 30mA limit, or was it just the best that could be done with someone's design?

 

[idleprocess]

Re: Introducing an EDC 2-Cell LED Light (CAD) SPY0

I'm wondering how much machine time will be required for that design. The shape strongly suggests injection-molding. It looks like about 80% of the blank will be swept away as chips.

A conventional 3-axis mill would require working the piece from 4 different sides. Use of a rotary table or some very granular CNC-control of X-Y to work the curved profile would be necessary.

If you've got the machine, and can dedicate the time, that's great...

 

[McGizmo]

Re: Introducing an EDC 2-Cell LED Light (CAD) SPY0

Data point to Newbie's comment, I recently changed the low sense on a light I EDC (Wiz2x2) and it now is set at 18 mA and works just great! Since the high is only 300 mA, I wanted a greater spread.

 

[Data]

Re: Introducing an EDC 2-Cell LED Light (CAD) SPY0

idleprocess

No less than two machine shops will be trained on these parts. My machine is a full four axis mill. It is clear to me from your comments you know something about machining. I agree with you that the SPY will take considerable machining time. That is why I am planing for a December roll-out.

/ubbthreads/images/graemlins/santa.gif /ubbthreads/images/graemlins/santa.gif /ubbthreads/images/graemlins/santa.gif /ubbthreads/images/graemlins/santa.gif /ubbthreads/images/graemlins/santa.gif

Cheers
Dave

 

[idleprocess]

Re: Introducing an EDC 2-Cell LED Light (CAD) SPY0

Ah, I wish I knew more about machining. Both hands gripping the rests on the armchair, I'm afraid. I used to do engineering work, but that was almost entirely sheet metal with a sprinkling of molded/formed plastic parts and a few machined parts - mostly flat stock that was machined into a bezel of sorts or slotted to act as a guide.

I'm familiar with how a 3-axis machine works as well as some of its limitations. There was an old Bridgeport manually-operated mill where I used to work - about all it did was mill slots the punch wasn't up to.

I should really take a class or something on machining - it really interests me, but living in an apartment is a bit of a show-stopper on setting up a shop...

 

[Anglepoise]

Re: Introducing an EDC 2-Cell LED Light (CAD) SPY0

Dave,
Great design and I wish you all the success in the world.
The only comment I feel qualified to make is in regard to 'O' rings. As mentioned by others, metric 'O' rings are very useful when you have size restraints. 1mm cross section can do wonders.
Now my main concern is with the switch 'O' ring.
Static rings give no problems but when you have movement involved......now you have the potential for torque problems. 'O' rings that are required to move will have different torque resistance as the lube dries out.
Starts off smooth as silk and 10 days later you can not believe you are operating the same switch. Get this right as the switch is a vital part of the customer experience.

 

[Data]

Re: Introducing an EDC 2-Cell LED Light (CAD) SPY0

Anglepoise

Your assessment is very astute and very appreciated. It is good food for thought.
/ubbthreads/images/graemlins/bowdown.gif

I have to agree with you. If this o-ring was sandwiched between two standard anodized surfaces it would eventually be ground to bits. Anodizing is very hard and on a micro scale, some what like sand paper. There is no way lube is going to stay between this o-ring and the knob. As you say, in so many turns the lube will be pushed out.

I do however have a perspective on this and here it goes.

As you know the body has a groove cut just for the o-ring to set into. It is hard anodized aluminum and therefor not the slipperiest surface around. When the o-ring is placed in there it will be touching a good amount of surface area and will lock in place. So on this side of the o-ring there will be no wear.

The knob is polished to a "mirror" surface even on the back edge where this o-ring touches it. And that mirror polished surface is going to be chrome plated. Chrome is very hard and somewhat slippery. Now add to all this either a Buna-N or polyurethane o-ring. Either of which is rated with an excellent wear rating. I plan to only crush the o-ring about 0.007", I think this will be just right and I believe the o-ring will last indefinitely.

Due to your and others concerns on this detail I will be testing this design with hours of dry friction to see exactly what it can take. Consider that this is precisely the kind of dry seal configuration used in many rotary devices that do not and can not be lubricated.

Now, having said all that I welcome your perspective as to how best handle this detail. I myself prefer polyurethane for this o-ring but it sounds like you know this stuff really well, what do you think would be the best material and configuration?


Cheers
Dave

 

[jdriller]

Re: Introducing an EDC 2-Cell LED Light (CAD) SPY0

I like it!

 

[Data]

Re: Introducing an EDC 2-Cell LED Light (CAD) SPY0

McGizmo

Here are four images of the LED mounting plate and the body where it fits in.

Sorry, I forgot to model the wire holes in your little PCB that comes with the McCapsule. I am going to need a bunch of those!

The four lugs in the body that the plate sets up against are minimal in size and may have to be moved because they encroach on Wayne's territory. We have not reviewed there exact placement yet.

Cheers
Dave

 

[Data]

Re: Introducing an EDC 2-Cell Light Feeler(CAD) SP

[ QUOTE ]
McGizmo said:
Dave,
I took another look at your renderings to see if I could make some other heads that would exchange with the stock one and. . .



[/ QUOTE ]

Please explain what you were asking for again. I was not following.

Dave