Need help building a wirelessly powered engagement ring

[KillerSpud]

Hi everybody!
So I have a really cool idea for an engagement ring that seems to me to be entirely possible (if not especially efficient).  I want to build a ring that has LEDs embedded inside.  Obviously there is very little room for a battery, so building an inductive coil seems to be the obvious choice.  I figure the power supply can live in the ring box so as to light it when placed in the box, but also removable so it can light the ring while being held in the hand.  I also think getting white LEDs would be best, so I need to source some very small warm white SMD LEDs.

I've done a good amount of thinking on the design, but I'm not the best when it comes to magnetics.  What I have planned on doing so far is to solder magnet wire to the ends of SMD LEDs to form the ring.  I am debating whether I should connect the LED rings in series or in parallel.  I think I can get away with only one or two turns per LED, and I am confident enough in my soldering skills to not make too much of a mess of it.  I bought a pack of wire from radioshack of 22, 26 and 30 AWG wire.  I'm thinking the current running through the ring is going to be pretty small so the 30 AWG wire should work just fine.

Now the part that I'm really not sure about is the power supply.  The designs I've seen online are just transistors and instructors flapping back and forth to generate the AC.  But being a uC guy, I kinda prefer the idea of driving it with a Tiny13.  I kinda figure I can use a regular 2n2222 connected directly to the batteries and coil and rely on the internal resistance of the batteries (probably AAA or coin cell) to limit the current, and implementing some duty cycle regulation in the uC to keep things from catching fire.  I also plan to use the larger 22 gauge wire on this side for greater chooch factor.

The one part I have no idea how to approach is encapsulating the ring.  I would like some kind of opaque resin plastic, but it has to be tough enough to be worn at least for a while, and be able to be cast around the electronics.  (I plan to use some kind of internal ring structure to wind the wires around)  Any help or direction here would be greatly appreciated.

So, what do you guys think?  Is there anything I need to keep in mind to make this work well?  Is there a better way than just two loops when they won't be places side by side (this is why I figure its going to be very inefficient).  I guess the biggest question is how I should build the driver circuit?

--Greg
(sorry for the wall-o-text)

**EDIT: See my update post on page 2.

[CJay]

It's quite astounding how little power is needed to produce visible light from a modern LED, there's a long running thread over on another forum about an LED from a party balloon using two LR41s in July 2015 and is still lit, albeit very faintly, as of 26th November 2016 on the same LR41 cells.

So, with that in mind, three possible sources of power spring to mind which might be worth investigating.

Light, Heat and RF

A single, non phosphor LED will work as a current source when illuminated by ambient light, it may be possible to use that to charge some energy storage device, maybe capacitors.

Hopefully your prospective partner will be warm, thermocouples might be useable, they produce very low voltages but at very low source impedance, might be possible to series connect a number to produce some workable voltage from body heat, ditto Peltier junctions.

It's possible to read non battery UHF RFID tags at ranges of 15 feet, ultra long range RFID is a thing, again you'd need to 'harvest' the energy and store it until you have enough to flash the LEDs. May also be possible to harvest enough energy from local RF to flash the LEDs occasionally and as most people these days have cellphones it could be workable to harvest some of that RF.

[TiN]

Or find some paint/material which emits visible light under IR light, so you can make it glow from external IR light source, which can be embedded on tie or other part on clothing Sell both pieces as a kit and you are in business

[KillerSpud]

Short of building a very LARGE coil system, like the size of a table lamp or bigger, you're not likely to be able to transmit several milliwatts of power effectively to it more than about a foot away.
Honestly I think for this to be very "practical" you'd want it to, say, work almost anywhere in a small to medium sized room.   Maybe within a 3 meter radius.  That is just not going to happen with any kind of simple engineering and transmitter / receiver.

I actually never intended to get more than a couple inches of range out of this.  I was thinking that the power source would be in the ring box with the ring, or held in the same hand that the ring is worn on.  That would limit how often it would be on, but I think that would be sufficient.  The LEDs I found (the 1206 white ones on sparkfun.com) need about 64mw, what kind of power would I need to push to light that up at a distance of 3-4cm?

[alanb]

I don't think that it will ever replace Diamond!

[grifftech]

copper and zinc on the inside

[KillerSpud]

Killerspud- it's been done previously! See if you can build on what was done here.

http://www.kokes.net/projectlonghaul/projectlonghaul.htm

That's a cool project and almost exactly what I want to do, except I'm not going to use the metal band, and I want to make the power supply much smaller.  But thank you for the link.  It has some good info in it.

[daveshah]

The LEDs I found (the 1206 white ones on sparkfun.com) need about 64mw, what kind of power would I need to push to light that up at a distance of 3-4cm?

That's probably a maximum power, if you wanted a dim but visible glow even a hundredth of that might be acceptable.

[mikeselectricstuff]

Electrically this should be pretty easy. I'd use a coil of very fine magnet wire (way thinner than 30AWG) with as many turns as you can manage, to maximise voltage. You want to maximise the coil area so should be wound "around the finger"
I'd guess 50-200 turns. tune this with a capacitor to form a parallel resonant circuit, and white LED(s) in parallel across it. Not sure if there would be any benefit of putting half the LEDs in anti-parallel to catch both half-cycles. The series R of the fine wire should keep the Q low enough that the resonance peak is wide enough that you don't need a super-precise excitation frequency.
I'd look at somewhere in the 100-500KHz range. Maybe 125KHz would be good so it would light when placed near a Low Frequency RFID reader. 
The coil would resonate up to the Vf of the LEDs, which would then start conducting and clamp the coil voltage to 3V

LEDs are not a big problem - you can get 0402 warm white LEDs on Aliexpress

The killer thing mechanically is that the ring can't be made of metal as this would severely damp the field. Bear in mind that even non-magnetic materials will block the field due to eddy-current losses.
It might be feasible to do it with a gap in the metal so it doesn't form a complete ring electrically, and use a metal with low resistivity

You might get slightly better performance with a rectifying diode and cap, though probably not enough   space to have a cap big enough to get any signifcant afterglow when removed from the transmitter.

[Marco]

These would fit in a chunky ring.

[mikeselectricstuff]

This might be another option for small power storage - 7 dia x2.5 thick
http://www.vishay.com/capacitors/aluminum/energy-storage-smd/

You'd need a voltage booster, so woul dneed to watch quiescent current issues. Should be doable within the same diameter as the cap

Nice thing is that being a cap, charging is easy. However inductive charging would be harder due to the battery itself ( not sure if these are magnetic)
Maybe a small ferrite cored inductor, but would need to be lined up with charger coil - maybe a special stand.

[jonovid]

induction loop system may just work?  that one hot dress you're wearing.
I had this idea for blinking leds on magnetic whiteboard markers using induction loop type frame.
but a ring, maybe out of range of any induction loop, unless its at a very high power level. 
why would you microwave the bride, in the name of science. 

[joseph nicholas]

How about a gps tracking engagement ring for those with suspicious minds?

[Marco]

Could a couple of chip inductors like this be used as a receiving coil?

[mikeselectricstuff]

Could a couple of chip inductors like this be used as a receiving coil?

interesting - I'd not noticed ship coils with inductance that high before - Certainly worth  a try but I suspect the loop area might be a bit low.

[nctnico]

In my experience with wireless power I don't think it is trivial to come up with a good scheme. At some point it will need regulation at either the receiver or transmitter side.

[Marco]

In my experience with wireless power I don't think it is trivial to come up with a good scheme. At some point it will need regulation at either the receiver or transmitter side.

At the receiver you could do something simple like having an IR LED and turning it on or off when peak voltage gets too high. That would give the transmitter enough to work with I think.

[mikeselectricstuff]

The trick here is that a white LED will produce light over a wide range of power - 100uA to 20mA. The internal resistance of the receive inductor will limit current, so no need for regulation.
The trick is to figure out the right balance of inductor inductance  - higher = higher voltage, but thinner wire - lower current.

[nctnico]

The trick here is that a white LED will produce light over a wide range of power - 100uA to 20mA. The internal resistance of the receive inductor will limit current, so no need for regulation.

It is not that easy. The voltage  depends on the distance so you basically will get an impedance mismatch. To get a reasonable transfer over a bit of distance (say a few cm) you'll need to use resonators because pure magnetic coupling won't work. This means you can't load the receiving resonance circuit too much otherwise the Q factor will be too low. But it all depends on how well you can control the distance, power levels and resonance frequency.
Also safety may be a concern if the transmitter is capable of outputting a few Watt. Any metal in the field can heat up to high temperatures quickly.

I'm surprised till now no one has ever posted this: http://www.kokes.net/projectlonghaul/projectlonghaul.htm .

Because it is an utter waste of time from a romantic perspective! Better knock your girlfriend up and say 'let's get married' to her when she tells you she is pregnant.

[mikeselectricstuff]

The trick here is that a white LED will produce light over a wide range of power - 100uA to 20mA. The internal resistance of the receive inductor will limit current, so no need for regulation.

It is not that easy. The voltage  depends on the distance so you basically will get an impedance mismatch. To get a reasonable transfer over a bit of distance (say a few cm) you'll need to use resonators because pure magnetic coupling won't work. This means you can't load the receiving resonance circuit too much otherwise the Q factor will be too low. But it all depends on how well you can control the distance, power levels and resonance frequency.

We're not bothered about efficiency - we just want to light a few LEDs. Even a few % will work fine. Small size is te goal here.

Also safety may be a concern if the transmitter is capable of outputting a few Watt. Any metal in the field can heat up to high temperatures quickly.

This shouldn't need enough power that heating is an issue - we only need a few mW max for the LEDs so even at a few % efficiency, input power should be well below anything that could cause heating

[KillerSpud]

Hello again and thanks for all the info.  I was successful in building a proof of concept:

All the basics are there, I just have to build a more realistic version to make sure it will work in the space and size I'm aiming for.