Calling all electronics geeks

[athon]

I'm in need of your skills. I'm wondering how one might go about making a simple, solar powered LED light that will work (even if just for a short while) away from a light source, such as the sun. Emphasis is on the word 'simple' - it needs to be done with a minimum number of components, as cheap as possible, and be something a child could do.

Ideas?

Athon

[alfaniner]

http://www.instructables.com
has tons of stuff like this, with detailed instructions.

[Dilb]

Absolute simplest way would probably be a solar cell hooked up to a capacitor, resistor and LED. I forget if solar cells allow current to flow through them in the dark, so you might need to connect in in parallel, with one branch being the capacitor, one branch being the LED and resistor. Make sure the RC time constant is a few seconds. I'm pretty sure I built this once using a kid's electronics kit.

[athon]

http://www.instructables.com
has tons of stuff like this, with detailed instructions.

Yeah, I use Instructables all the time. Great site. However the closest I can get to what I want on there is still either a little too complicated or simply suggests getting something that already does the job and modify it. Since this is for an activity, I want to know if there's a really simple way of accomplishing it that doesn't require a heap of work.

Absolute simplest way would probably be a solar cell hooked up to a capacitor, resistor and LED. I forget if solar cells allow current to flow through them in the dark, so you might need to connect in in parallel, with one branch being the capacitor, one branch being the LED and resistor. Make sure the RC time constant is a few seconds. I'm pretty sure I built this once using a kid's electronics kit.

Ok, this is more what I was thinking. I was considering using a capacitor, but my experience with electronics is pretty minimal (I was far more into dissecting animals than electronics as a kid :p).

Any safety hints? I know it might seem like a daft question, but in my line of work I'm always over cautious.

I'll give it a shot.

Athon

[geni]

You need a solar cell. A resistor. A capacitor and an LED.

http://www.technologystudent.com/elec1/capac1.htm

[geni]

Something like this with the solar cell as the power source:

[geni]

Any safety hints? I know it might seem like a daft question, but in my line of work I'm always over cautious.


Throw enough power at an LED and they explode. It's not a vast amount of power so googles are recomended unless you feel like digging bits of LED out of kids eyes.

[athon]

Throw enough power at an LED and they explode. It's not a vast amount of power so googles are recomended unless you feel like digging bits of LED out of kids eyes.

Ok, cool. Didn't know that (I would have thought they'd just melt).

Anything else a novice like me needs to know about capacitors? What size would be appropriate? (yes, I realise it's probably a 'how long is a piece of string' question...)

Athon

[CaveDave]

Ok, cool. Didn't know that (I would have thought they'd just melt).

Anything else a novice like me needs to know about capacitors? What size would be appropriate? (yes, I realise it's probably a 'how long is a piece of string' question...)

Athon

Depending on the size of the PV cell and the LED used, I would start around 10 microfarad. With a single PV cell (or two in series) a 5 volt rating would be enough. If you use a 12v or larger PV module, I'd use a capacitor of twice the voltage rating or more to give some headroom.

The resistor should be sized to limit peak current to the LED according to it's specs. See here (http://en.wikipedia.org/wiki/LED_circuit) for how to calculate it, but often 100 - 1000 ohms is enough (wattage is low, so 1/8 - 1/4 watt is probably big enough)

Low power LEDs will draw less, so they will "hold over" longer in darkness, all else being equal. Red/green/yellow types are easier to power than blue, IIRC.

HTH

Dave

[Ivor the Engineer]

ETA: Read the following with reference to the circuit diagram in Geni's post.

Ok, cool. Didn't know that (I would have thought they'd just melt).

Anything else a novice like me needs to know about capacitors?

Yep, the larger values (approx. 1uF and larger) are typically polarised and have a maximum voltage rating which can be quite low (6V). Connecting them the wrong way or to too higher voltage supply can cause them to explode.

Aluminium capacitors are clearly labelled with their value, voltage rating and polarity. To get the lead which should be either positive or negative with respect to the other lead, look for either a string of '+''s on one side of the capacitor (more positive lead) or '-''s (more negative lead).

My finger in the air guess is you will require a 470uF to 2200uF capacitor to keep an LED on long enough to use as a useful example of the principles involved. The voltage rating will depend on the maximum voltage from the solar cells, but 35V capacitors are inexpensive and likely to be more than sufficient for typical solar cells.

One final word of warning: The value of capacitors you will probably be using for this experiment can store significant amounts of charge. Shorting their leads together when charged to several volts will likely produce a small spark, and possibly cause the leads to weld themselves together.

Having said all that, probably the most likely injury is scratched/pierced fingers from handling the small and quite sharp component leads. Have a supply of plasters to hand!

What size would be appropriate? (yes, I realise it's probably a 'how long is a piece of string' question...)

Athon

Not at all. How long do you wish the LED to remain on after switching the lights out?

For a rough order of magnitude estimate, a couple of simple relationships can be used to size the capacitor:

Q = Charge in Coulombs; C = Capacitance in Farads; V = Voltage in Volts; I = Current in Amps; t = time in seconds.

Q = C.V - (1)

Q = I.t - (2)

Equating (1) and (2):

C.V = I.t

Rearranging:

C = I.t / V - (3)

Using the initial conditions of V, the voltage supplied by the solar cell under the specific lighting conditions just before the lights are turned off, I, the current drawn by the resistor and LED in series with V volts across them and t, the desired time for the LED to stay on after the lights are turned off.

Example:

Say the solar cell produces 6V under the lighting conditions used for the experiment, the resistor+LED is to draw 1mA with 6V across them and it is desired for the LED to remain lit for 2 seconds after the lights go out:

C = 1mA x 2s / 6V = 333uF.

So values from 470uF to 1000uF should keep the LED lit for approximately 2 seconds after the lights go out. Capacitors of this value will almost certainly be polarised.

To calculate the value of the resistor required to limit the current to 1mA with 6V across it and the LED in series, use ohm's law:

V = I.R

The forward voltage of the LED needs to be figured into the calculation. Hight Efficiency (HE) Red LED's typically have a forward voltage of about 2V, so the resistor value to limit the current to 1mA is given by:

R = (Vsupply_max - Vled) / Imax = (6V - 2V) / 1mA = 4k-ohm. Nearest preferred value = 3.9k-ohm.

In practice it may be better to allow a slightly higher current in the LED to give a brighter light initially.

As with all these things, experiment with different values of resistor and capacitor until you have a reliable experiment for others to perform.

[wackyvorlon]

Throw enough power at an LED and they explode. It's not a vast amount of power so googles are recomended unless you feel like digging bits of LED out of kids eyes.

Don't think I've ever seen that one. Most of the time, they get hot and let the magic smoke out. I've never had to wear goggles in dealing with LEDs, and some are designed for a surprising amount of power. Failure of the LED is prevented by building in some current limiting, usually something like a 330 ohm resistor.

[geni]

Don't think I've ever seen that one. Most of the time, they get hot and let the magic smoke out. I've never had to wear goggles in dealing with LEDs, and some are designed for a surprising amount of power. Failure of the LED is prevented by building in some current limiting, usually something like a 330 ohm resistor.

Yes but you are an adult. Kids may well connect the LED dirrectly to the power source.

[shadron]

Ok, this is more what I was thinking. I was considering using a capacitor, but my experience with electronics is pretty minimal (I was far more into dissecting animals than electronics as a kid :p).

Any safety hints? I know it might seem like a daft question, but in my line of work I'm always over cautious.

I'll give it a shot.

Athon

In a resistor/capacitor circuit (in fact, for all passive circuits - those involving only resistors, capacitors and inductors), the voltage across the capacitor will never exceed the voltage that the solar cell puts out. So in your circuit above, it will never exceed the voltage output by the solar cell.

This means that there is no danger of shocking the user. However, it is still possible to have a fire or burning hazard by, say, shorting out the capacitor's terminals with steel wool or crossing it with a finger ring, which could heat up a lot in the current flow. It is essentially the same problem you have while working with a car's 12 volt battery - no shock, but don't short out the terminals. Remove all hand and arm jewelry before working on it. Keep away from steel wool and flammables.

[shadron]

Don't think I've ever seen that one. Most of the time, they get hot and let the magic smoke out. I've never had to wear goggles in dealing with LEDs, and some are designed for a surprising amount of power. Failure of the LED is prevented by building in some current limiting, usually something like a 330 ohm resistor.

It depends on how well sealed the LED is, and how fast it heats up. It can be explosive if well sealed and it heats up fast.

[lightfire22000]

Well, do you need to make it from scratch, or are you willing to use a kit?

[Dr. Trintignant]

So values from 470uF to 1000uF should keep the LED lit for approximately 2 seconds after the lights go out. Capacitors of this value will almost certainly be polarised.

You can get much larger capacitors nowadays for a low price. For instance:
http://www.mouser.com/Search/ProductDetail.aspx?qs=sGAEpiMZZMuDCPMZUZ%252bYl40X %2fGhnxjhBQQvmYzOu2tc%3d

0.22F = 220,000 uF. That's a very useful amount of energy for this purpose, and only about a dollar. The 5.5v rating should be sufficient. Or:
http://www.mouser.com/Search/ProductDetail.aspx?qs=sGAEpiMZZMuDCPMZUZ%252bYl40X %2fGhnxjhBkfoJ454hUuo%3d

A full farad for <$3.

- Dr. Trintignant

[athon]

Fantastic. Thanks once again everybody, especially Ivor for those figures.

I still have a few more details to sort to see if this could make for a potential activity (that pun was quite unintentional). But it gives me a starting point to do an activity on capacitors and solar cells.

Cheers.

Athon

[PbFoot]

Don't think I've ever seen that one. Most of the time, they get hot and let the magic smoke out. I've never had to wear goggles in dealing with LEDs, and some are designed for a surprising amount of power. Failure of the LED is prevented by building in some current limiting, usually something like a 330 ohm resistor.

I've managed to explode a few LEDs by connecting them directly to a fresh 9V battery. The smell is terrible. You'd need a lot of solar cells and some very bright light to pop an LED though.

-PbFoot

[lightfire22000]

If you're going to make a lot of things, you might want to try making your own capacitors and inductors.

[athon]

If you're going to make a lot of things, you might want to try making your own capacitors and inductors.

I'm always open for new ideas. Feel free to add them here, or pop them into a PM and send it to me.

Athon

[Dilb]

If you're going to make a lot of things, you might want to try making your own capacitors and inductors.

For power applications, it would be really difficult to build your own capacitors. Home made capacitors are usually huge, and stilil only have capacitances in the nF range. A Leyden jar charged up to a few kV could technically store enough power, but obviously that's a terrible design to run an LED.

[Ivor the Engineer]

You can get much larger capacitors nowadays for a low price. For instance:
http://www.mouser.com/Search/ProductDetail.aspx?qs=sGAEpiMZZMuDCPMZUZ%252bYl40X %2fGhnxjhBQQvmYzOu2tc%3d

0.22F = 220,000 uF. That's a very useful amount of energy for this purpose, and only about a dollar. The 5.5v rating should be sufficient. Or:
http://www.mouser.com/Search/ProductDetail.aspx?qs=sGAEpiMZZMuDCPMZUZ%252bYl40X %2fGhnxjhBkfoJ454hUuo%3d

A full farad for <$3.

- Dr. Trintignant

Good idea. I'd forgotten about supercapacitors.

One other thing to mention is a reverse-blocking diode might be required between the solar cell and the capacitor. This is because solar cells can leak significant amounts of current when a voltage is applied across their terminals in the dark. To avoid wasting power, a low forward voltage drop diode such as a 1N5817 should be inserted with its anode connected to the positive output of the solar cell and its cathode (lead nearest stripe on body of the diode) to the positive of the circuit being powered.

[athon]

Ok...

Sorry to resurrect this thread, but I've finally had a moment to dig through my store cupboard and see what I have handy.

I've got a 2.0 volt solar cell, a 1000uf 16V capacitor and a bag of yellow LEDs. Given this, what should I buy from the electronics store? (From what I've seen around, 2.0 volt photovoltaic cells are the best compromise for cost - I'm not keen on getting kids to fork out more than $20 per cell).

(sorry for seeming a tad slow on this topic - while I'm familiar with the concepts and terminology, in practice I've never been one to dabble in electronics).

Athon

[Dilb]

2.0 volts isn't enough from a solar cell, a yellow LED needs to have at least that before any current flows. The extra voltage will be what actually causes the LED to light up. You could put two, or preferably more, solar cells in series to up the voltage.

If you have two solar cells in series, you have 4 volts, but only 2 of those volts will be across the resistor. For 1 mA of current, you want
R=V/I = 2V/1mA = 2000 Ohms
and this has a time constant of
t=R*C = 2kOhm*1 mF = 2 seconds
so the brightness will drop to 37% in 2 seconds. Less resistance will be brighter, but decay faster, more resistance will be dimmer but will last longer.

You might also want a 3 state switch (either bought or built yourself), so you can change the capacitor between two circuits: one with just the solar cells to charge the capacitor, an open circuit where the capacitor stays charged, and the LED and resistor. That way you don't need to worry about the capacitor discharging through the solar cell. Otherwise you need a diode like Ivor said, to keep current from going back through the solar cell.

[athon]

Ah, sod. That makes sense.

Hm, ok, thanks for that info, Dilb. Exactly what I needed to know.

I know it seems petty, but the nature of these activities demands cheapness and simplicity. I can occasionally make room to budget a bit more, but if the components get too expensive, I start excluding a lot more kids/parents/schools.

Athon

[Dr. Trintignant]

Ah, sod. That makes sense.

Hm, ok, thanks for that info, Dilb. Exactly what I needed to know.

I know it seems petty, but the nature of these activities demands cheapness and simplicity. I can occasionally make room to budget a bit more, but if the components get too expensive, I start excluding a lot more kids/parents/schools.

Athon

Do you have a target cost in mind? You said above that >$20 is bad, but what do you think is reasonable?

I did a Google search and found this:
http://www.hobbyengineering.com/H2202.html

6.7v, 30 ma. That's enough for two LEDs in series and has a very decent current rating (or, use one LED and have a longer runtime). $12/each, less in quantity. The store has another model with half the current rating (still enough for an LED) for $9.

I suspect that you could provide all the parts for somewhere in the $10-$15 range, depending on how many kits you bought and what kind of performance you wanted.

- Dr. Trintignant

[Ivor the Engineer]

A three-state switch is probably a better idea than the blocking diode.

The actual current rating of the solar cell is not particularly important if it only has to charge a capacitor which is then switched over to power the LED, so the 20mA panel from the link Dr. Trintignant provided would be more than adequate.

As others have said, what you need to find is a solar cell with sufficient voltage to charge a capacitor to a higher voltage (say >5V) than the LED's forward voltage (around 2V).

[Dan O.]

You can use a circuit like the Joule Thief (http://www.evilmadscientist.com/article.php/joulethief) to charge a battery from a solar cell or drive an LED from a single battery. If you scrounge the parts from old electronics it's almost free.

[wackyvorlon]

I'm on a QRP ham radio mailing list, and a while back one of the members built a transmitter using almost entirely parts scavenged from a dead compact fluorescent bulb:) He had to add an inductor, as I recall.

[athon]

Do you have a target cost in mind? You said above that >$20 is bad, but what do you think is reasonable?

Good question. There really isn't an ideal range; it's simply a matter of 'cheaper, simpler, better' when it comes to designing or finding activities.

The aim of the newsletter I write is to encourage as many kids as possible to explore some aspect of science, which means aiming for those who would have nothing more than what you'd find in the average home. Of course, this isn't a strict rule and I try to find ways of pushing the limits on this occasionally.

I did a Google search and found this:
http://www.hobbyengineering.com/H2202.html

Oh the fun I could have if I could source all of the materials online. It'd open a world of possibilities.

At this point, I restrict myself to things that can be found easily in common local stores. I will stretch that to hobby stores on occasion, such as this one.

There's a few reasons for this, and it is complicated. For one thing, Australia hasn't quite caught up to the rest of the civilised world when it comes to finding the internet useful (my wife - an American - still finds it most frustrating). So the surrounding culture makes it difficult. Another reason is that kids generally don't have credit cards or the means to order stuff easily online without their parents' help, and although I'm aware that the newsletter is filtered through adults in 90% of cases, intentionally requiring their direct involvement too much is something I avoid.

6.7v, 30 ma. That's enough for two LEDs in series and has a very decent current rating (or, use one LED and have a longer runtime). $12/each, less in quantity. The store has another model with half the current rating (still enough for an LED) for $9.

I suspect that you could provide all the parts for somewhere in the $10-$15 range, depending on how many kits you bought and what kind of performance you wanted.

- Dr. Trintignant

I'll keep hunting locally, but might have to store this whole idea away for another situation. I still want to do something with capacitors and solar cells, but might need to rethink how I do this.

Most helpful, though. Thanks heaps!

Athon

[athon]

You can use a circuit like the Joule Thief (http://www.evilmadscientist.com/article.php/joulethief) to charge a battery from a solar cell or drive an LED from a single battery. If you scrounge the parts from old electronics it's almost free.

That's cool.

It's making me think there's a niche market for a newsletter like mine, only aimed specifically at promoting science technology amongst kids who want to learn more about electronics. I keep overstretching my boundaries, unfortunately, although I think it's important for kids to get into electronics from a young age.

Athon

[Dr. Trintignant]

I'll keep hunting locally, but might have to store this whole idea away for another situation. I still want to do something with capacitors and solar cells, but might need to rethink how I do this.

Most helpful, though. Thanks heaps!

Athon

Glad to help. For some reason, I had thought you were a science teacher or perhaps something like a scout leader, where you would be giving instruction in person. In that case, it would be a bit easier to deal with the online aspect since you could do all the ordering yourself. A newsletter approach definitely makes things harder.

I really like wackyvorlon's suggestion of utilizing parts inside a compact fluorescent bulb. They're filled with lots of nifty parts, and only cost a few dollars each. The problem is that disassembly is hard, especially with the safety considerations around the glass bulb. So that's probably a non-starter.

Ever thought about putting together packs of electronics components that you could resell to the kids? Probably more work than you're willing to commit to, but it sure would make things easier.

- Dr. Trintignant

[athon]

Ever thought about putting together packs of electronics components that you could resell to the kids? Probably more work than you're willing to commit to, but it sure would make things easier.

The market is essentially cornered by Dick Smith electronics (same Australian skeptic who Randi inspired to start the Australian Skeptics). We do have a range of science kits, and it might be worth looking into one day, but it's not really my field.

Athon

[wackyvorlon]

The aim of the newsletter I write is to encourage as many kids as possible to explore some aspect of science, which means aiming for those who would have nothing more than what you'd find in the average home. Of course, this isn't a strict rule and I try to find ways of pushing the limits on this occasionally.


Are your readers in the US? If I may, I would like to submit Radio Shack as a potential source. They've really beefed up their selection in recent years. I would recommend possibly considering a circuit called the atari punk console. Linky:

http://www.getlofi.com/?p=518

It can be made from parts readily available at radioshack, and makes some delightful bleeps and boops. Adafruit Industries markets some small kits that are very nifty, as well.