Make them glow forever!
February 23, 2008 12:33 PM Subscribe
I want to make LED Throwies, but I want them to last much, much longer than 2 weeks. I'm thinking 6 months or more. Is such a thing possible? How would I do it? What kind of batteries would I use? I don't know much about electronics, but I understand the basic equations. Thanks!
Perhaps rather than soldering the LED directly to the battery, you'd solder it to a battery clip and replace the batteries as needed.
Those throwies are using 3 V LEDs. If you can stand a larger throwie, you could use a bracket that holds a pair of AAA batteries, and that way could even use rechargeable batteries (and probably get much longer burn time as well). Radio Shack or Fry's will have this.
posted by adamrice at 1:03 PM on February 23, 2008
Those throwies are using 3 V LEDs. If you can stand a larger throwie, you could use a bracket that holds a pair of AAA batteries, and that way could even use rechargeable batteries (and probably get much longer burn time as well). Radio Shack or Fry's will have this.
posted by adamrice at 1:03 PM on February 23, 2008
A 3V CR2032 battery has a capacity of around 230mAh, but it can only be discharged at a few mA, meaning that it can last anywhere from 75 to 230 hours. Six months is about 4300 hours, so for the same current levels you'll want a battery with 4300 to 12900mAh. It'll be bigger and more expensive than the tiny batteries. Keep in mind also that larger batteries will often discharge at higher currents, so you may want to put in a current-limiting resistor in series with the LED.
posted by Krrrlson at 1:06 PM on February 23, 2008
posted by Krrrlson at 1:06 PM on February 23, 2008
LEDs draw about 20milli-amps (mA) at their brightest. Battery capacity is rated in milli-amp-hours (mAh). So you can divide the capacity by the LED current draw to get lifetime of the battery. The Wikipedia List of battery sizes is pretty helpful here. The typical CR2032 battery (often used in throwies) is shown as having 225 mAh capacity, which means you will get about 225 mAh/20mA = 11.25 hours.
A few things you can do to extend this is to put a resistor in between the battery and LED to restrict the current. But this also makes the LED dimmer. You can maybe double a battery's lifetime without detracting too much from the brightness with this trick.
The other direction is to use bigger batteries. If you used two D cells (19500 mAh for alkaline) in series, you get 19500 mAh / 20 mA = 975 hours = 40 weeks. You'd definitely need a resistor here though, because D cells have very low intrinsic resistance. LEDs need their current restricted or they blow up. A 200 ohm resistor should suite you well. Oh and the reason why you need two D cells in series is you need greater than 2 V to turn on red or green LEDs, ~3V to turn on white & blue LEDs.
Of course D cells are pretty bulky and that throwie turns into a ballistic weapon. :) To get back to being smaller, you need to use tricks like pulse-width modulation. PWM turns the LED on & off really quickly and uses our human persistence of vision to make it look like it's on all the time. This requires a bit of engineering and extra circuitry beyond the typical throwie circuit of just LED & battery. Do a search on instructables & google for "programmable LED" and "smart LED" for one approach to this. (disclaimer: I have a little product called BlinkM that is an implementation of a smart LED)
posted by todbot at 1:11 PM on February 23, 2008
A few things you can do to extend this is to put a resistor in between the battery and LED to restrict the current. But this also makes the LED dimmer. You can maybe double a battery's lifetime without detracting too much from the brightness with this trick.
The other direction is to use bigger batteries. If you used two D cells (19500 mAh for alkaline) in series, you get 19500 mAh / 20 mA = 975 hours = 40 weeks. You'd definitely need a resistor here though, because D cells have very low intrinsic resistance. LEDs need their current restricted or they blow up. A 200 ohm resistor should suite you well. Oh and the reason why you need two D cells in series is you need greater than 2 V to turn on red or green LEDs, ~3V to turn on white & blue LEDs.
Of course D cells are pretty bulky and that throwie turns into a ballistic weapon. :) To get back to being smaller, you need to use tricks like pulse-width modulation. PWM turns the LED on & off really quickly and uses our human persistence of vision to make it look like it's on all the time. This requires a bit of engineering and extra circuitry beyond the typical throwie circuit of just LED & battery. Do a search on instructables & google for "programmable LED" and "smart LED" for one approach to this. (disclaimer: I have a little product called BlinkM that is an implementation of a smart LED)
posted by todbot at 1:11 PM on February 23, 2008
The 11 hour figure doesn't account for the low discharge rate of Lithium cells, todbot -- I think that's how they were able to get a week out of them. Also, I'm not sure how PWM will help, since the switching losses are likely to consume more power than they save, and the extra ICs will make the cost prohibitive.
Forgot one thing: if you should ever need to use those current-limiting resistors, here is a handy utility.
posted by Krrrlson at 1:16 PM on February 23, 2008
Forgot one thing: if you should ever need to use those current-limiting resistors, here is a handy utility.
posted by Krrrlson at 1:16 PM on February 23, 2008
Perhaps you could put double their lifetime by putting in circuitry to turn the led off during daylight. You could even generalize this a bit more by only turning the led on a few hours after sunset and an hour before sunrise on the theory that the throwie will have no audience at 2:00 am.
posted by rdr at 1:31 PM on February 23, 2008
posted by rdr at 1:31 PM on February 23, 2008
Todbot and Krrrlson have pretty much covered everything I'd add about battery issues. I like rdr's suggestion, and I think you could do the "turn off during daylight" thing pretty easily with a photoresistor and a transistor (adding maybe $1 to the cost of each throwie). The question with some of these additions is: does the added cost, size and assembly time outweigh or overshadow the simplicity and cheapness of the original design that makes throwing a couple of hundred of them out into a fun and affordable idea? It depends on the application -- if you still want to throw a pile of them onto metal objects in the neighborhood, increasing the cost beyond a certain threshhold ruins the economics. If you're building a small number of "throwie-like" devices for some other purpose, then added cost/complexity is okay..
posted by Alterscape at 1:51 PM on February 23, 2008
posted by Alterscape at 1:51 PM on February 23, 2008
You know, I bet you could do something like this in some locations (like Chicago), but it would be very technically demanding.
I'm thinking of a throwie which would have an attached pinwheel generator designed to be driven by wind, and which would recharge the battery. A generator requires a conductor moving in a (non-uniform) magnetic field, and you already have quite a strong magnetic field from the rare earth magnet used to make the throwie stick. All you have to do is come up with a rotor with a bunch of fine wire loops in it, which would attach to the shaft of the pinwheel and be rotated in the magnetic field whenever sufficient wind came up. The current generated would have to be rectified and controlled so as to be able to charge the battery, but I think there are plenty of circuit designs out there which can do that.
Actually, the most direct way to do it would be to mount the LEDs, coils, and all circuitry and the battery to the spinnable pinwheel and have the whole assembly rotate freely with respect to the base containing the magnet. The spinning LEDs might produce interesting novel visual elements (color mixing, etc.), and you could even have a design with no battery which would only light up when the wind blew.
Hmmm, I think a wind-spinnable throwie would be interesting in its own right.
posted by jamjam at 2:45 PM on February 23, 2008
I'm thinking of a throwie which would have an attached pinwheel generator designed to be driven by wind, and which would recharge the battery. A generator requires a conductor moving in a (non-uniform) magnetic field, and you already have quite a strong magnetic field from the rare earth magnet used to make the throwie stick. All you have to do is come up with a rotor with a bunch of fine wire loops in it, which would attach to the shaft of the pinwheel and be rotated in the magnetic field whenever sufficient wind came up. The current generated would have to be rectified and controlled so as to be able to charge the battery, but I think there are plenty of circuit designs out there which can do that.
Actually, the most direct way to do it would be to mount the LEDs, coils, and all circuitry and the battery to the spinnable pinwheel and have the whole assembly rotate freely with respect to the base containing the magnet. The spinning LEDs might produce interesting novel visual elements (color mixing, etc.), and you could even have a design with no battery which would only light up when the wind blew.
Hmmm, I think a wind-spinnable throwie would be interesting in its own right.
posted by jamjam at 2:45 PM on February 23, 2008
How about some sort of induction loop which could pick up power from power lines? It would limit the possible locations, but...
posted by alexei at 5:39 PM on February 23, 2008
posted by alexei at 5:39 PM on February 23, 2008
an idea: check out the datasheets for blinking leds. my guess would be that if they are off half the time, they would use less current. (of course, they may just waste the current as heat.)
also, this is a cool circuit for a throwie that only turns on at night. if you're putting these outside, then you can double the life of each throwie.
posted by kamelhoecker at 5:53 PM on February 23, 2008 [1 favorite]
also, this is a cool circuit for a throwie that only turns on at night. if you're putting these outside, then you can double the life of each throwie.
posted by kamelhoecker at 5:53 PM on February 23, 2008 [1 favorite]
Although they're hard to find now, the LM3909 chip could give a good flash out of an LED for months on a C cell. It consumes a tiny amount of current and there are replacement circuits available - more work for a throwie, but doable.
posted by plinth at 6:15 PM on February 23, 2008
posted by plinth at 6:15 PM on February 23, 2008
« Older life is a song that begins with the word behold | You want to see this e-mail with attachment? Ok... Newer »
This thread is closed to new comments.
posted by atrazine at 12:43 PM on February 23, 2008