InfraRed Power Transmission

(for wireless cellphone charging)

​

The source paper describes R&D work targeted at sending modest amounts of power (< 1 watt) on a narrow� infrared beam over distances up to about 25 feet.

​

The approach and results look promising. This is not quite ready for a product yet because there are pieces of the total “product” that still have to be developed.

​

Let’s take this in 2 pieces:

​

1. What the reported work actually accomplished

2. How this might become a real product – what is possible, what are the drawbacks

​

​

In the slides to follow, please forgive my very amateurish drawings, this is not one of my strengths

​

The authors have developed a sophisticated system for generating a very columnated infrared beam:

< ½” diameter over at least 15 feet is necessary

​

​

Not to scale:

​

​

​

Transmitter,�Control

InfraRed�Beam

Cell Phone

Slightly Reflective�Hemispheric Lens

How it works:

​

1. Transmitter sends out narrow beam at very low power

​

• Hemispheric lens assures that angle of incidence is not critical

​

• A small amount of signal gets reflected back from lens

​

• Control system senses reflected signal, knows signal is reaching the phone

​

• Transmitter turns on full power

​

• If anything intersects beam, control system detects this and immediately cuts back to low power

​

• Light is focused by lens onto photovoltaic cell in phone which converts light to electric power

​

The actual “handshake” between units and transmission of power was demonstrated

Issues and TBD

​

1. Transmitter and phone must be “line of sight” – phone cannot be in your pocket or purse

​

• Lens must be facing transmitter

​

• Possible design: Put transmitter on the ceiling, more or less centered. Place phone on any surface, lens side up

​

• “Search mode” for beam to find phone is not described.

​

• Possible problem: Some reflective surface that, by coincidence, reflects beam back to transmitter when control system is “looking for” phone

​

• Possible solution” Replace simple reflective lens with coded low power IR transmitter inside phone that sends signal out through the lens. Then by (pulse) coding the transmitted signal and having control unit look for “the right response” proper identification is guaranteed. Variations of this could also encode signal strength information to help control unit find the phone.

​

• Various “smart” designs for having the units find each other are easy to come up with. The difficulty will be the required beam width and steering controls needed in the control unit.

​

This is a good piece of work – it’s much easier to narrowly focus a light beam than to focus a radio wave beam. The issues scale with the wavelength of the beam. These authors have provided a necessary piece of the puzzle.

Repeat of the False Positive statistics calculation – Just in case it went by too fast in the video

​

Suppose we have a test for some disease that has two problems:

​

1. 1% False positive – If I test 1,000 people that DO NOT have the disease, I’ll get 10 positive results

​

• 2% False negative – If I test 1,000 people that DO have the disease, I’ll get 20 negative results

​

​

You take the test. We’ll look at the implications of you testing positive or testing negative separately below.

​

​

What is not immediately obvious is that the validity of the results depends on how prevalent the disease really is.

​

Suppose the statistics of the disease are that 1 person in 1,000 in the population (0.1%) have the disease.

​

​

​

​

Note: The calculations to follow are approximate, albeit pretty good. I’ll show why at the end

Case 1: You test positive. What’s the probability you have the disease?

​

If we test 1,000 people, on the average

1 person actually has the disease

10 people get false positive results

​

The probability the you actually have the disease is 1/11 ~ 9%

​

​

Case 2: You test negative

​

If we test 1,000 people on the average

1 person actually has the disease

​

The probability that we get a false negative is .02

​

Therefore, the probability that you don’t have the disease is 98%.

​

​

​

Error: Adding 1 + 10 above isn’t exact, they overlap. The actual total is something like 10.9