Featured Post

Theranos CEO Elizabeth Holmes Finally Faces Criminal Charges

It's been some time coming, but the CEO of Theranos is finally facing criminal charges for fraud, as the WSJ's Carreyrou repor...

Monday, July 30, 2018

More Clarity from New Energous FOIA Documents

One of the pleasures of writing this blog is in the communications with engineers and scientists who are experts in their respective fields, and I get to learn from private email discussions, or have a great sounding board to bounce theories back and forth with. Following my recent posts on the Energous FOIA documents, one contributing reader let me know they had also directly requested those documents on Energous, and received a batch of redacted emails that had some new ones not in the collection that had been previously available, which led to some interesting exchanges - so thank you for that.

These documents cover an earlier time period from December 2014 to March 2017, with at least one of the email threads having a full year gap in the middle. Those of you who would like to see the new documents in full, email me and I'll send them, or you can contact the FCC yourself and say "I ask for the responsive records for FOIA 2018-000342", and start with the FCC FOIA page here.

There are no stunning technical revelations in here, but there are some key points from earlier that seem to be confirmed. For those who don't want to read the details, the main takeaways are that:
  • Switch from 5.75 GHz to 915 MHz seems to have happened late summer 2016
  • The "local" RF power issue to get Part 18 seems to be resolved in Oct/Nov 2016
  • Reduction of power at receiver to well below 1 Watt seems to happen at end of 2016
  • Explicit admission of single device charging occurs in early 2017

These four items combined are likely to have a significant impact on the practicality of the device - huge "pockets" of energy and poor steering, highly limited operating range, delivered power below useful levels for most devices, and only one device at any time. The entire system was on shaky ground to begin with, IMO, and these changes made what I would consider a terrible product far, far worse. In future posts I may compare the dates of these changes, that have quite an impact, with statements from Energous at the time. A publicly traded company must be very careful in what it tells the public.

"Throw it at the wall and see what sticks"
I guessed that Energous were just refiling applications again and again, each time making what minimal changes they thought might get them through, and relying on the FCC to guide them in changes, or perhaps simply to give up and so win through attrition. These documents seem to reinforce that with five face-to-face meetings between June 2016 and March 2017, and nine OET Submissions between August 2016 and March 2017. If we look at the times they submitted, there are some key changes at various dates. Unlike most situations where a company has a product that has to meet minimum specs to satisfy customer demand, it's obvious that there is no aspect of the design that will not be sacrificed in order to get Part 18 approval, no matter how pointless the resulting 'product' is. I list the physical meetings and OET submissions at the base of this article. Key changes are:
  • 5.75 GHz to 915 MHz change occurred between June 23rd 2016 (experimental licence) and October 17th 2016 (Document 59)
  • Part 18 "Unlimited Power" appears to have been resolved between October 17th 2016 (Doc 59) and November 8th 2016 (Doc 58), possibly at the October 26th face-to-face meeting. "Local" and "unconstrained" RF energy issues do not arise anymore. Note this does not resolve safety issues, such as SAR
  • Two Receivers Down to One happens between August 23rd 2016 (Doc 61) and December 12th 2016 (Doc 56). No reason given, perhaps complex SAR measurements
  • Drop Below 1 Watt Charging Claim happens between August 23rd 2016 (Doc 61) and (likely) December 12th 2016 (Doc 56) or (definitively) March 28th 2017 (Doc 5)
  • Change from 10 Antenna to 12 happens between December 12th 2016 (Doc 56) and February 26th 2017 (Doc 1)
  • Change from Sound Bar to Angled Sound Bar happens between December 12th 2016 (Doc 56) and February 26th 2017 (Doc 1). May just be a description change, but coincides with # of antenna change. Energous still demonstrating 'abandoned' straight bar as of January and June 2017.
  • Change to Charging a Single Device is in February 26th 2017 (Doc 1), although it was obvious even in earlier submissions this was likely. May be a move to "time sharing" charging multiple devices, reducing already low charge rate to each

No significant changes seem to have occurred between March 2017 and approval in December 2017, which surprised me as the system and performance were so basic that I thought they had to have rushed it, but it seems they had most of a year.

Given the above, it doesn't appear that the change to 915 MHz, at least by itself, fixed the Part 18 issues as that change had occurred by October 17th 2016, while Part 18 questions were still being asked. It could be that the change occurred and it just took time to get that accepted as sufficient for Part 18, or it could also mean that they still had not satisfied SAR safety and did not start lowering the power output until December 2016/March 2017. This also means that Energous knew the mid-range transmitter would be incompatible with the higher frequency mini-WattUp, or have 1 Watt charging, while promoting licencing deals with the likes of Myant. I wonder if Myant knew?

The most important change that allowed for Part 18 approval seems confirmed now to be the requirement for "local" RF energy, and that such energy is never "uncontained". I've suggested this may restrict any such system from working in the far field (which for an array this size is around 1 meter at 915 MHz), but may also be restricting the system from working in the near field where there are multiple "pockets of energy". This would then be a second need for the safety cutoff system that prevents SAR limits being exceeded. If this is the case, even without SAR limits the system would be constrained to the 50 cm to 1 m range it currently is, and makes questionable the claims of Energous' CEO that power limits could be raised by extending this keep-out zone.

This gives a possible explanation as to why 5.75 GHz was not used (beyond the simple FCC statement of "no destroying WiFi"), as the far-field boundary moves much further out and the "pocket" of energy gets smaller in theory - that any close in operation that would be needed would be in the near field with many maxima and minima, with potential for further maxima beyond the charging location. Either that or such a system would require too many antenna and electronics too precise for Energous to want/be able to build. There's still not enough information to resolve this question, but more pieces are beginning to fall into place.

If that's the case, it doesn't bode too well for Ossia, who recently claimed a shift to 5.8 GHz. While they have a 2D array and can probably dynamically alter the transmit aperture, say to a 30x30 cm square for a 1 meter far field. There would be a pretty reasonable number of elements in that, assuming 1/3 wavelength pitch (around 17 by 17, or ~300). Quite how they overcome the SAR and safety issue and still get reasonable energy out, I'm not sure - I still only see this working for IoT devices at exceptionally low power. Searching on the FCC website though, I can't find grantee code for Ossia - it's as if they've never interacted with the FCC on product regulation at all. Does anyone know what their FCC grantee code is?

So the outcome of this is more clarity that Energous were simply doing whatever it takes to get Part 18 approval, even if it were a product that failed to meet their original claims, and some visibility into the methods they used to slowly reduce capability, or wear down the FCC, until it was finally allowed. The "local" RF energy question seems to have been answered by placing severe restrictions on the usable range, while safety was met by reducing power again and again, retesting until it finally passed. Little by little, we're learning more about how this process was playing out in the background. I'm looking forward to the next set of revelations to narrow down what's really going on.

Oh, and Energous earnings tomorrow - odds on this ~$400m market cap company earning more than $25,000 this quarter? (My mistake, earnings at the end of Wednesday, not Tuesday!)

(Repeating the seemingly obligatory statement - I have no financial position, short or long, in Energous or any related company)

Below is a simply a summary of meetings and submissions to the FCC by Energous on their mid-range device. This is not intended, or likely, to be complete, just what is known at this time. I'll update with other information later.

Physical Meetings
  • 8/9th June 2016 (experimental licence application)
  • 27th July 2016 (Doc 61)
  • 26th October 2016 (Doc 55)
  • 6th December 2016 (Doc 54)
  • 28th Feb 2017 (Doc 3)

Office Engineering and Technology (OET) Response Submissions
  • 23rd August 2016 (Doc 61) - Still not clear that Part 18 is achievable, and claims a second receiver at 1 Watt. To get Part 18, repeatedly references "local" RF energy. Two receivers listed, one at 100 mW, one at 1 Watt - last mention of two receivers or 1 Watt. Power reported as "Number Receivers Supported: Rx-1: 100mW, Rx-2: 1W", No indication if these two receivers were on the same device, or could be separate, I expect a single device. Lists a sensor needed for SAR compliance.
  • 17th October 2016 (Doc 59) - Earliest mention of the 913 MHz band in available documents. Ongoing discussion of viability of Part 18 with it clearly critical that energy be assured to be "local" and can never be "uncontained". Also see page 2
  • 8th November 2016 (Doc 58) - All Part 18 discussion is dropped following Oct 26th meeting. It is never raised again in future documents. Part 18 "local" and "unconstrained" issues solved?
  • 30th November 2016 (Doc 55, Page 4) refers to an OET Response on this date but not available
  • 10th December 2016 (Doc 57) refers to meeting of Oct 26 and not Dec 6. Appears Energous want OET to develop new safety criteria for them. Jeff McNeil of Energous adds "Regulatory" to his SVP Ops title.
  • 12th December 2016 (Doc 56) refers to Dec 6th meeting. FCC requests clarity on which of the many designs will be submitted. Table 1 lists 10 antenna and "Target Platforms: Sound bar for desktop usage". Power now reported as "Cumulative Receive Power at 30 cm and 1m: [Redacted]". Use case claims "charging multiple devices", but admits to one device at a time.
  • 18th December 2016 (Doc 55) FCC request a different method of calculating SAR and measure individual antenna
  • 26th February 2017 (Doc 1) Table 1 lists 12 antenna and "Target Platforms: Angled Sound Bar for desktop usage". If an actual change, a limit on phase array capability to focus? Number of antenna and description change. Use case now admits "charging single device"
  • 28th March 2017 (Doc 5) - FCC still asking for single clear submission

Experimental Licence Application
  • To demonstrate the technology to FCC, a 15 day experimental licence was applied for , starting June 8th 2016 for 15 days. Application is here. The details of this can be found at the FCC site here, and clearly show still operating at 5.8 GHz, with 20.4 Watts output (55.6 W ERP). Energous' next such application was for 6 days, between 6th and 12th January this year for  CES, which listed the 913 MHz band with 30 W ERP. 

Thursday, July 19, 2018

Energous FCC FOIA Docs

I had a chance to go through the FCC/Energous FOIA documents mentioned in my last blog post. They span a period from around February to December 2017 when Energous were trying to get their mid-range transmitter approved under Part 18 (unlimited power) rules. Most of them have the really useful information redacted, and are Energous bugging the FCC to "please can we visit so we can move this on quickly" and the FCC engineers saying "errr, sorry out of town that day" like an ex-girlfriend trying to avoid an annoying and slightly psycho ex. One of them was super interesting with lots of info in it though, which reveals a lot of the history of the system.

The data shown seems to be for the system as they were trying to push through in early 2017, with 21 W out, but to ensure they hit the SAR safety limit by December had to dial it back to 10 Watts out. This ties with the December device power at 90 cm being 30 mW, while in March it was 60 mW at 1 meter. They pushed the size of the safety cutoff zone up to 50 cm from around 35 cm, possibly changing the focal point along the way (just eyeballing the structure it looks to have changed between March and December). 

The main FCC concerns seemed to be safety via the SAR limit, as well as ensuring that energy was in "pockets of concentration" (Doc 18). There's a significant discussion on corrective factors applied as safety margins, and basically they have to scale all measured results by ~1.5 and still be under the 1.6W/kg limit (Doc 30). This means, as I suspected, that Energous cannot raise the power output of their system from where they are today (0.966 W/kg) - what they have now is as powerful as it gets.

My read of this - Energous just kept resubmitting and resubmitting, each time with the FCC telling them to go away and what to do next, and eventually they dialed everything back to the point where they got it through under the SAR limit. An undercurrent in the notes is sloppy work by Energous, with the FCC constantly having to ask for clarifications, updates, or actually doing calculations for them! (Doc 30) 

I don't see any evidence of pressure on the staff to push the device through, more exasperation on their part with second rate engineers just throwing stuff at the wall hoping it would eventually stick.

The rest of this analysis is a little dry for those not interested in the details, just warning you.

There's a few things though that stand out as important, beyond what I listed earlier. First of all, they had moved to 915 MHz from 5.8 GHz by February 2017. Document 1 indicates that the February submission is a second clarification or change in response to a meeting they had with the FCC in October 2016. I could speculate that a 5 or 6 month response time indicates that some significant changes had been made, as new measurements or clarifications could be made quickly. This may be the timeframe in which the frequency switch was made. This IMO is a significant change with implications for performance of the system, yet was never mentioned in SEC calls or quarterly reports.

The work presented by Energous also appears to be sloppy, with the FCC multiple times noting how poor the data consistency and quality is (Doc 9 "trying to understand field distribution", Doc 41 requesting "proper and consistent information", Doc 48 "We understand that it was prepared quickly, but we suggest paying attention to some details.").  Their two tables don't match, for example the received power listed below when converted from dBm in Table 2 are 512 mW and 47 mW (27.1 and 16.7 dBm) for 30 cm and 1 meter respectively, but in Table 1 (previous post) are 190 mW and 60 mW (22.8 and 17.7 dBm).

Basically, the numbers don't add up. I expect the numbers here are for an earlier system that was very tightly focused to try and maximize power at a single point to get to ~500 mW to charge a phone, but ended up going way over SAR. There may be some other reason, one being estimated and one measured, but I'm tending to the "sloppy" for now.

Document 32 shows the "keep out" zone changed around November 2017, increasing it to 50 cm. This may also be when the physical structure of the power bar was changed, or may simply be that they were forced to update their SAR measurements and this was required.

Document 18 references the need not only to be concerned with safety and states that data "should also show that the energy distribution through field maps demonstrate that there are pockets of concentration". This indicates that the safety restriction was not only a single SAR number, but the physical distribution of the energy. 

Figure 11 above shows a typical on axis beam, with a peak at one point showing the transition from near to far field, and then a gradual but continuous reduction with distance. If we assume a "pocket" of energy implies a region where there is a lower value of energy both before and after the charge location, then that would restrict the use of the system to just beyond near field only. Even with a phased array, where that transition zone is, it's a function of frequency and physical size of the transmitter. For any practical size transmitter, (equations here) it's basically likely to never be viable beyond 1 meter. If this is the case, that's a huge limitation for any at-distance wireless power system unless they make the transmitter the size of a wall.

Also, that peak in Figure 11 is at 42cm, but they say a focus at 65cm. They might want to take a look at their work there...

The size of a region of constructive interference, a focus, is often defined by the half power, or -3dB points. Looking at Figures 11 and 13, it seems that region is about 60 cm in the x- and z-axes, but only 15cm or so in the y-axis. Some "pocket"!

The "safety system" that detects if anyone is in range of the device and shuts it off (supposedly) is ultrasonic, using TI PGA450 chips and my best guess would be Murata ultrasound transmitters (sound familiar?), used in car parking sensors. I'd be very interested to see how this is setup, because as someone who has worked with them before and ultrasound a lot, I think that system might be easy to fool if you don't design it very carefully.

So overall this data is interesting, nothing too amazing, but confirms what was suspected - that this system is at the limit of what it can transmit safely, that it doesn't have enough power to charge a phone in any realistic way, that efficiency is low, and that FCC staff weren't too impressed with Energous' consistency and quality of work. But it does prove that persistence pays off.

(Seemingly obligatory statement - I have no financial position, short or long, in Energous or any related company)

Energous: It's Worse Than We Thought

A contributor to SeekingAlpha has recently posted an article highlighting data from an FOIA request to the FCC regarding Energous' approval for their mid-range transmitter last December. They appear to show concern from the FCC officials as to how the system was performing, and understanding the system behavior. There's also discussion regarding the safety limits, which I think are one of the key issues and concerns surrounding this approval, and covered them in several posts. The author has provided the FOIA documents for download here.

For some background, I wrote several posts on the FCC approval in December last year, you can find them starting here.

Interestingly, earlier this year I did a FOIA request for all of FCC Chairman Ajit Pai's communications regarding Energous (along with several other keywords) and was told there was nothing. Given Ajit Pai's (IMO clearly illegal) use of the official FCC communications to promote a private company, I suspect there are documents there, just not available. I may revisit that.

One of the key images that has been redacted from the FCC report, but is floating around from another source, shows the actual specifications for their desktop system.

There are some key points in here:
  • The system works for only 1 receiver at a time
  • Output is from 12 antenna, each antenna is 1.8W (32.5 dBm) - 21.6 Watts total output
  • Receive power at 30cm is 190 mW, at 1m is 60mW
  • Max range 1 meter
  • Receiver is 6.5 cm in diameter
  • No mention of safety limit distance
There is no clear detail is that received power is actual RF power at target, or power to battery at target - I suspect the former. This is a higher output power than the system shown in FCC documents in the Part 18 Approval (21.6W vs 10W), and the receive power slightly lower - I was estimating 100 to 150 mW at 50 cm, some were estimating higher. That implies a "Wall to Battery" efficiency of 0.2% at 30cm and 0.06% at 1 meter, assuming 90 Watts in at the wall socket (thanks to a reader for pointing out my initial mistake here). That would mean a phone would take over a day to charge at 30 centimeters, and nearly 4 days at 1 meter - and that's assuming 100% efficiency on receive, and I also suspect those numbers are ideal and real world will be worse. You might say it's more appropriate for IoT or small item charging, however the receiver, at 6.5cm diameter, is wider than my phone. I expect it's multiple dipole antenna and they need it that size to get even that terrible efficiency. I can see why Myant didn't want this receiver in their underwear.

So charge times are obscenely long, it's incredibly inefficient, only one target receiver at a time, the receiver is larger than a phone, and it needs a safety cutoff if you get too close. Did I miss anything? 

Apple must be chomping at the bit to get hold of this technology...

I'll dig into the released documents in more detail later, but at first glance it doesn't quite match with prior statements from the Energous CEO as to system capabilities.

"Here is a brief summary of the results of the amount of actual power delivered to a device at varying distances with a single WattUp transmitter. Power received at zero to five feet measured 5.55 watts compared to our targeted performance of 4 watts. Power received at five to 10 feet measured 3.74 watts compared to our targeted performance of 2 watts and power received at 10 to 15 feet measured 1.06 watts compared to our targeted performance of 1 watt."

5.5 Watts vs an actual 0.19W - only a factor of around 29. Remember that when viewing the statements from at-distance wireless power companies as to their performance specs, compared to what they have to write in the legally required documents and spec sheets.

There's an update to this post, reading the releases in more detail, here.

(Since I always seem to need say this - I have no financial position, short or long, in Energous or any related company)