Can I power an LED from a cellphone's microwave output?

Question

I want to build something which picks up the microwaves from a cellphone and converts it into DC current to power an LED. I have a full wave rectifier; how should my antenna work?

Answer 1

You could build a circuit that amplfies the incoming RF power so that if any RF power is present, the amplified signal will light the LED.  The problem is that, in this day and age, RF energy is almost ALWAYS present!  And it's unclear what good such a device would do you anyway.  Even if you could tune it to the exact frequency band of the cell phone you're interested in, and limit it so that it only activates for THAT phone, cell phones transmit all the time, even when you're not using them.  They are in constant communication with the cellular system to manage power levels and hand-offs between cell sites.  

Answer 2

Do a basic ballpark calulation of power (assuming all your matching is perfect and no losses):

Led can use about 1.5V at 15mA for a decent amount of light, which equates to 22.5mW.  Your mobile phone (I assume) will have a max output of 100mW.  So in order to power the LED in the ideal case, you will need to obtain ~25% of the power generated by the antenna.

If you consider the phone antenna as an isotropic radiator (which isn't the case but is a close enough approximation), you can see that because it isnt a very directional antenna, you will need to occupy much of the space around the phone with a high gain antenna (in addition to being right next to it) in order to get that kind of signal.

That is the best case and is very difficult to accomplish, I would say requires way too much engineering to accomplish and may be impossible due to losses.  Your phone doesnt radiate in a continuous fashion anyway so the LED would blink very fast (maybe like a PWM driven LED) plus its power can vary.

If you do find a way around all this, your rectifier will need to be a fast switching type with RF diodes.  The LED just uses too much power in my opinion.

You need to go active in order to make this do-able, that's what I reckon anyway.  If you do engineer a way around it then I would be interested to know how.

Answer 3

Basically you need to make a detector circuit, something like a crystal radio. The first link below explains this. You need some selectivity - a tuning circuit - that will pick up the bands of interest (900 Mhz and 1.8/1.9GHz for a cell phone).

Diode detectors always have a sensitivity issue: you need enough voltage to overcome the diode's Vf. For a Schottky diode Vf is as low as 300mV, a regular silicon diode it's about 600mV, and for an LED it's much higher: 1.2V or more depending on the diode type.

In fact, your full-wave solution makes things worse since it adds two diode drops plus that of the LED, so your detector wouldn't trigger unless you had 2.4V available. The LED alone in half-wave would perform better. This might be good enough if, say, you want the detector right on the phone.

How to overcome this? The trick is to add a DC bias to overcome Vf so that the small RF voltage will kick the voltage over the Vf hump to light the diode. I realize this pours cold water on the idea of your detector being self-powered, but it will work better than a passive solution.

I think a better circuit would use a Schottky diode for detecting then buffer the output to drive the LED. Even better than that, I've also linked an IC that can do this. It's very low current and can be powered by a Li coin cell. It's also very tiny.