Project Lamplighter - my homemade LED keychain fob

[AnAppleSnail]

It'll be visible from as many angles as I can manage. It's going to be partially made of stainless steel or brass or copper or titanium (or whatever looks the coolest). I'm having it CNC'd, which means this is going to cost a fortune - I hope it's worth it. Once it's made and I post all about it here, if you think it's worth it, I'll make you one too.

Count me as 'interested,' and eager to see what I can get in exchange for mere monies.

 

[calipsoii]

Count me as 'interested,' and eager to see what I can get in exchange for mere monies.

I've been working pretty diligently on this project so I'm getting cautiously optimistic that I could have a prototype soon. How's that for beating around the bush? :huh:

Seriously though, the CAD's are complete and the dimensioned drawings are done. All the other parts have been ordered, save for the PCB's, which are designed and just need printed at the factory.

Since I have absolutely 0 engineering/machining experience, right now I'm just talking to machinists and finding out how thin you can machine different metals, what some standard threading sizes are, those kind of things. I'm concerned that if they try and machine what I have drawn, it'll just explode on the lathe or something.

:duck:

Need to get a quote too - hopefully this won't break the bank too badly.

 

[derfyled]

Definitely too quiet...

How is the runtime test, still glowing ?

 

[calipsoii]

Definitely too quiet...

How is the runtime test, still glowing ?

Still glowing!

I wish I could say that I have a prototype to show off, but not quite yet. I've been so busy lately that I just haven't had the time. It kills me to have this project sit stagnant but right now it's best to deal with everything else that's going on and come back to this when I can do it properly.

 

[3zuli]

great idea! I made one on Sunday (after reading this thread), just a red smd led connected directly to a LR44 (1.5V) button cell battery. it glows! let's see, how long it lasts.

I want to try this flasher: http://www.discovercircuits.com/DJ-Circuits/OSC4.htm (the one on the right)

so it seems that it has died... last time I checked it was about 2 weeks ago, in the night you could still notice the weak glow. the single LR44 battery has leaked in January (iirc), but it kept glowing even after that
still, that's almost 6 months of glowing. 6 months = 162 days, so let's assume it's been glowing for 150 days. 150 days = 3600 hours, LR44 has about 150mAh capacity. 0.15Ah/3600h = 0.0000416A = 0.0416mA average current.
now, let's make one with C or D 1.5V battery, that should last for years :devil: Alkaline D has 12000-18000mAh (wikipedia), so let's say 15Ah/0.0000416A = 360 576.923h and that is 41.16 years. Let's do this

btw, this looks interesting too: http://translate.google.com/transla...u=http://hawelson.blog.cz/0701/vecna-svitilna

 

[calipsoii]

Looks like the fat lady has sung. A little over 6 months of 24hr-a-day runtime isn't so bad though. :devil:

I've been getting quotes from machine shops but so far the cheapest I've found has a minimum of 100 pieces and is sitting well outside my comfort range to pay out of pocket.

He did say that the 100 pieces could be done in any kind of metal though: stainless steel, aluminum, brass, copper or titanium (for a higher price). :naughty:

I'm going to keep looking and I'll let you know what I find!

 

[calipsoii]

This entire assembly would fit comfortably inside a 10mm LED (with room to spare) :devil:

 

[Dr.Jones]

I just took a look at the circuit - and I think it has a much lower efficiency than you think.

calipsoii wrote:
Homemade constant-current circuit. If my calculations are correct, it's driving the LED at 0.005ma
The batteries are driven in series, making them capable of delivering 6V, but they really only need to provide 1.1V to meet the forward voltage of the LED. The circuit is supposed to be a very efficient direct-driven one, so I can't imagine it's worse than 80%. So how long will it run for? I'll tell you when this little guy goes out.


Well, it's a constant current regulator, so it's by definition NOT direct drive. Also it's a linear driver, so it's efficiency is input voltage divided by output voltage - at maximum.

V+ ---+-----+
......|.....|
......R2...LED
......|.....|
......+-----T1
......|.....|
......T2----+
......|.....|
......|.....R1
......|.....|
GND---+-----+

I assume you use two standard NPN transistors, since there is no need for a FET here. Also from the images I read R1=R2=100k.

If the LED needs 1.2V and the battery gives 6V, then only 1.2V/6V=20% of the power in the LED-T1-R1 branch is used for the LED.
The current in that branch is about 0.6V/R1 = 6uA (like what you calculated)
However, the base of T1 is at 1.2V above GND, so R2 drops 4.8V, which takes a current of 4.8V/R2=48uA through the R2/T2 branch.
Thus only 6uA/54uA=11% of the total current goes through the LED.
Both together result in about 2% total efficiency, if I didn't miss something.

Possible solutions:
To increase "voltage" efficiency, use 3 cells or use 2 LEDs. Runtime will decrease *a bit* though.
To increase "current" efficiency, use a much higher R2, like 1M, 2M2 or 4M7.

 

[CarpentryHero]

I hope that you do a run of these

 

[AnAppleSnail]

I've gotten the lamplighter bug again. Coming soon, I hope.

 

[AnAppleSnail]

I've gotten the lamplighter bug again. Coming soon, I hope.

 

[SoCalDep]

I just read the thread and what a cool idea! Very much paying attention to the progress!

 

[calipsoii]

I've gotten the lamplighter bug again. Coming soon, I hope.

I like you AnAppleSnail, you've been commenting on this thread since the beginning and you're good at giving that little nudge I need every once in a while. :laughing:

I haven't been updating the thread, but I've certainly spent a lot of time on this project lately.

In the last few weeks I have:

Received 1200 LR41 batteries

400 gold-plated springs

33 feet of Pyrex tubing

The Pyrex is a real pain in the butt. It doesn't "score and snap" like your standard glass tubing does, it shatters irregularly around the score line. Even a wet-cut tile saw chips the edges all to heck.

I'm currently looking for a glass shop in the city that will cut it to length for me so that I can just flame the edges nice and smooth after they're done.

Tough stuff though! I took my car keys to it for 10 minutes and it didn't leave a single scratch. I then threw that little piece at a concrete wall and it didn't chip or shatter.

So what's the hold up?!

Money

I found a local machine shop here that I just love. The owner is amazingly friendly and he is used to doing very small precision work (like watch parts). He does jewelry as a hobby so he found my design very interesting from that angle. He's willing to make the final product in any combination of brass, copper or stainless steel (or titanium for 2x the price). He offers all kinds of surface treatments and if I commit to the full order, he'll make me a few prototypes to test and tweak before the final run.

The problem is that his MOQ is 100 pieces (very reasonable in my opinion) and each completed Lamplighter costs $36CDN to machine.

To give some perspective to that: 2 chinese machine shops quoted me $2100 and $2350USD for 100 pieces. Both ended up yanking my chain for a few weeks, refusing to send prototypes, not honoring quotes they gave me only a few days earlier, and just generally making the entire experience a poor one. After a couple weeks of talking to them, I'd had enough! I'd rather chat with this guy locally and be able to drive over and see the results than send the money away and not even know whether the pieces will fit together when they arrive.

You can see how spending $3600 on 100 pieces of a product that I don't even know if I can sell is a pretty tough pill to swallow though, eh? I've thought about starting a feeler thread to gauge interest, but haven't had a chance to yet. Maybe I should post a couple renders and see what people think of the design...

Anyways, I've saved 70% of the amount so far, so we'll see how long it takes to hoard the rest. I'll be sure to update the thread when we get there!

Thanks for poking your head in though - sometimes I need a little nudge to keep things rolling.

 

[AnAppleSnail]

If you have a reasonably durable keychain attachment, put me down for 4 or $200 worth, whichever is more expensive. I'll see if I can find a dollar store pyrex bowl to play with diamond saws on. I know you'd be TERRIBLY excited to hand-dremel-cut 1000 pcs and then polish them by hand. Have you tried at a university lab for the pyrex cutting? They have neat saws there.

As far as machining goes - Hrm. I'm reminded of the bayonet-base 24v bulbs that look like Google Image search image. I picture that with a threaded battery compartment instead of bayonet base and a keychain lug. My understanding is that lamp bulb bases clamp/solder onto a ceramic base attached to the glass bulb. Is that neat enough for your purposes? Only you know. This might simplify the machining, but may not help the cost.

By the way - these could be a steampunk version of tritium-vial necklaces. In a personal mission, I've been looking for ways to make the smallest one possible (Large resistor, LED, and two tiny hearing-aid batteries) and a very small pulsing circuit - a 'breathing' pattern for a small, dim light. So far those have all had runtimes in hours, not in months.

 

[nein166]

I cut glass tubes at work for sight glasses and we use a cutter with a small diamond wheel that goes into the tube, not sure you'd find one that will fit in whatever the ID tube you have, makes a really clean break when you heat the tube and dip it in cool waterup to the stress point.

 

[calipsoii]

I cut glass tubes at work for sight glasses and we use a cutter with a small diamond wheel that goes into the tube, not sure you'd find one that will fit in whatever the ID tube you have, makes a really clean break when you heat the tube and dip it in cool waterup to the stress point.

Hmm, regular glass or tempered borosilicate (pyrex)?

I used a diamond cutter to score the tubing fairly deep, then heated it until it was glowing and dipped it in cold water and it didn't even budge.

The tempered Pyrex is made to withstand very high heat without fracturing so I don't know if the hot->cold trick works.

 

[^^Nova^^]

Wow, these are looking sweet. Having the renders here would be good.

How are you sealing the ends of the tube? Are the batteries going to be replaceable?

Cheers,
Nova

 

[nein166]

not really sure of the type, theres someting writen on the tubes I'll check, probably not pyrex though they like to buy the cheap stuff we use it on the vacuum side of a steam turbine so its usually not over 150F

 

[mcbrat]

anything ever happen with this?

 

[calipsoii]

anything ever happen with this?

No, not yet.

That's not to say I've been standing still. I've just been doing a lot of learning.