Saturday, April 16, 2016

Those Other Guys, Pt 1

Having made comment about others' concerns over tech startups like Theranos ability to fleece investors and the public, I thought I would make some comments regarding Energous, another wireless power company getting some press. I'll point out here that I have no inside knowledge of the workings of Energous, just that I've spent a lot of time in the last few years looking at the physics of transferring power via waves (acoustic and electromagnetic waves have a lot of similarities in the maths and fundamental behaviours, much of what I say below can be applied to ultrasound phased arrays), and how investors in tech companies think.

For the less technically minded among you, I'll summarise in the next part the key points from here, and you can just skip straight to part II.

Energous uses RF electromagnetic waves, similar to those used in the higher speed wifi signals, to transfer the energy. At the 5.8GHz they claim, wavelength is going to be around 5.2 cm (divide the speed of light at 3e8 m/s by the frequency of 5.8e9 Hz). Why is this important? Here's an image from their patent filings, with the power transmitter in the top left, labelled 706:


It's clear from the patent they intend to use a phased array, similar to what is done in radar, and what uBeam has claimed to do. Essentially, it's a large regular grid of transmitters, like a chessboard, and by sending the signals from each small piece at slightly different times, you can send a beam to a chosen location. An image below shows a beam from an array from a standard computer simulation package. This is a well known phenomena and has been for many years, there are few surprises.
Notice as well as the main lobe (big red bit where most of the energy is going), there are lots of little peaks going in other directions called side lobes. These are an inevitable consequence in any practical system, and involve energy wasted and going places you don't want. You can also get even worse lobes called 'grating lobes' if you don't put those transmitters close enough together, less than half a wavelength or around 2.5 cm. Let's assume, like any good engineer would if designing a phased array where you are sending power and don't want to cook anything off to the sides, that you go with 2.5 cm spacing.

Now Energous have been claiming 'hundreds of small elements' so in the great power analysis here let's assume around 500, or about a 22 by 22 grid, which at 2.5cm spacing is 55cm on each side. Wow, that's a big panel. Let's make it closer to the size of a speaker, which is what the press is describing here and make it to 100, that's a 25 by 25cm plate, still pretty sizeable. That's not out of scope - if you look at the scale of array that Ossia, also using RF to do charging, is showing in their patents. (Ossia work at half the frequency, so wavelength is 2x bigger).
Anyway, Energous claim to create 'pockets of 3D energy' which is marketing speak for 'beam'. No pocket of energy magically appears with none elsewhere, like a magnifying glass with sunlight, it concentrates more and more until your point focus is very intense. How small you can make that focus is related to the wavelength L (smaller wavelength for smaller focus) and transmitter size relative to wavelength D (bigger transmitter is smaller focus), and how far away you are trying to focus F (further away is larger focus). Generally, your focus (pocket of energy) is given by L*F/D. That's not exact, but is close enough for a blog, and the fact that the energy slowly dies off, like in the image below, would certainly be a safety concern. Not something I want to be near even away from that pocket.

So using the above numbers, at 5 meters range (just beyond the 15 feet they state), the focus is 0.05 * 5 / .25, or around 1 meter in size. That's a big 'pocket'. Even at 1 meter range that's still a 20 cm focus which is larger than any phone, and will catch your hand, head, and anything else nearby. In an IEEE Spectrum article they state the pocket is larger than the receiver, so not unbelievable. (Note to uBeam: Energous seem to know how to handle a skeptical technical press, despite similar incredulity from the experts)

The 'mini-transmitters' they talk about are ridiculous though - do the above calculation with a USB sized device. It's no wonder it would literally take years to charge a phone with it, even before you consider you need to reach a level of 0.5 to 1 Watt these days to trigger charging.

And then there's directivity where the amplitude falls off with angle as you move away from the beam pointing straight ahead - by the time you get to 60 degrees, you're at half-power, and down to 0 when pointing to the side. This simply doesn't mesh with Energous' claims of a '30 foot diameter bubble'. In summary - with a phased array it's easy to point straight ahead, and harder as you move to the side. This means it gets less efficient - a significant problem when imaging, a killer if you're transferring power.


I won't go into other numbers showing that their claims of power delivery capability are not just difficult but defying laws of physics - the only thing more ridiculous would be claiming that ultrasound could charge a device that's in your pocket. Anyway, you can find information demolishing Energous' claims here including mention of the regulatory issues they face with the FCC, though everyone seems to forget the FDA can choose to become involved in any radiation emitting device (and ultrasound is radiation, FYI).

Confused, bored, lost? No? Then you probably knew most of this anyway, are technically proficient, and were already skeptical of Energous. Yep? That's what 95% of people who read this will feel like, and switch off. And that's the point.

That's the tech side - next part, the business side and their novel approach to fundraising.

14 comments:

  1. This comment has been removed by the author.

    ReplyDelete
  2. Hello,

    Have been reading your posts after coming across some uBeam related news. Just wanted to point to a couple of things and see what your thoughts on this are?

    1) The IEEE spectrum article you linked to mentions a 'Bluetooth connectivity'. I can see the array being phased and therefore electronically steered to work around the 'Gain drop' you correctly point to for highly directional transmitters. If this is paired with phased array tech. at on the receiver (on the cell phone) which is claimed to use meta-material (another pseudo-cool buzz word, I know ...eye roll), then is it not possible to maintain a 'line of sigh' with the max. gain angle of the transmitter (or both transmitter and receiver?).

    2) Secondly you are right about the antenna beam spreading out to a much larger area than that required for a cell phone and the amount of energy being high in absolute terms. However as far as high water content human tissue (fat, muscle, skin) is concerned the skin depth for EM energy penetration is very small isn't it? (approximated as 1/sqrt(frequency) SI units..ref H.P Shwan,1971 ).

    The questions about efficiency are conveniently dodged so that is definitely suspect. Moreover even if there was electronic 'adaptive' steering control loop implemented, it won't help matters if there's a large number of passersby in the rated range of the transmitter. But it's still possible to see this working in sparse environments, no? or even with a sufficiently large number of transmitters, latching on dynamically to different cell phones in the vicinity. There's problems with other bodies blocking the energy but don't quite see the fundamental physics in-feasibility of the technology.

    ReplyDelete
    Replies
    1. I too am an rf engineer with specialization in antennas so let me take a crack at your questions.

      1.) First of all, I am skeptical about their claims regarding meta-materials. Yes, it has been demonstrated in the literature that using the odd material properties of meta materials can result in electrically small antennas, which would allow for a higher directivity array (if more elements are added), but I somehow doubt they are making any huge breakthroughs there; I just think they're brute forcing it and creating a large enough grid of elements to get the directivity they want.

      Secondly, yes, beam-steering can change the directionality of the pattern, but typically, the more extreme the deflection angle from boresight, the more the directivity suffers and the more grating lobes become a thing, which constrains the maximum angle the device can scan to without hitting the performance, but I'm not even sure all that matters. The reason is because given the range Energous specifies and the apparent size of the transmitter (estimated from pictures of it), the device that's being charged has to remain IN THE NEAR FIELD OF THE TRANSMITTER ANTENNAS. The calculation for this is done by considering the overall transmitter aperture size. What this means is that any array factor and friis transmission formula calculations go out the window because they require the observation point to be in the far-field. Either Energous has missed this or they're betting no one will notice, but this makes their tech a glorified near field wireless charging station.

      2.) It is true that human body absorption at 5.8 GHz is relatively low, but that doesn't matter if enough power is in play. I think the device's transmit power is in excess of 43W, which would make me wary about standing directly in front of it. I'm not typically worried about general RF exposure, but 43W seems way too high to just allow without further investigation. Granted, they claim that, if obstructed, the beam will deflect away from the obstruction and the phone or whatever will charging via multipath, but the problems I noted in answer 1.) arise under that condition (limited deflection and grating lobes).

      Now can this work? Yeah sure it can, it's not altogether unfeasible, it's just not economical (efficiency is 12% best case no ohmic heating or mismatch with perfect polarization and line of sight and it's a Tuesady in February) or guaranteed to be safe, and there's a great deal of them talking out of their a**es, but what else is new?

      Delete
  3. Oh and don't get me wrong, I'm hardly batting for Energous here. Just digging deeper into the specific engineering difficulties you have brought up. I'm an RF engineer myself and the details interest me a lot.

    ReplyDelete
  4. What if Energous has developed with its Tier-1 partner a technology that includes sensors on the final product which actually disables charging when a hand or any human body is coming close or is holding it or if it comes across the formed beam ? Wouldn't that let the final product go through FCC if they could prove the "beam" or "pocket", whatever we call it never reaches the human body ? That doesn't sound like impossible to do. Does it ?

    ReplyDelete
  5. This comment has been removed by the author.

    ReplyDelete
  6. Sounds like a disgruntled Ubeam employee that maybe then tried to get on at Energous and got turned down there as well to me

    ReplyDelete
    Replies
    1. Sorry to disappoint, but I have never applied to nor have I been interested in working for Energous.

      Disgruntled means 'angry or dissatisfied' so yes, I am a dissatisfied former uBeam employee. That doesn't make me wrong. Rather than just throwing out statements like that, care to point out flaws in any of my logic, facts, or arguments concerning uBeam, Theranos, or Energous?

      Delete
  7. This comment has been removed by a blog administrator.

    ReplyDelete
  8. This comment has been removed by a blog administrator.

    ReplyDelete
  9. This comment has been removed by a blog administrator.

    ReplyDelete
  10. This comment has been removed by a blog administrator.

    ReplyDelete