訂閱電子報，掌握最新科技與產業趨勢
ChargeEdge
2019/08/07
ChargEdge重啟無線電力傳輸
FREMONT, Calif., Aug. 6, 2019 /PRNewswire/ -- On July 25, 2019, ChargEdge, an innovative startup out of Silicon Valley, well-placed historically to "save wireless power transfer (WPT)," gave an invited seminar, courtesy of the IEEE Power Electronics Society, SF Bay Area Chapter, at the Texas Instruments auditorium in Santa Clara. A group of seasoned engineers and academicians huddled together with the CTO of ChargEdge, Sanjaya Maniktala, a highly-published author in the area of power conversion, to better understand the underlying reasons for the vast potential of WPT having remained unrealized to date. With the help of accessible math spreadsheets and simulations, occasionally referencing stunningly simple, but powerful scaling laws he had discovered and published via McGraw-Hill in 2014, Sanjaya explained that there existed an industry-wide disconnect with the basic physics underlying resonance, leading to mediocre implementations and "mysterious" cancellations. He described how ChargEdge, "knowing-it-all" beforehand, had charted a unique course through a minefield of performance and regulatory issues, very quietly, finally arriving with the full promise of true wireless power in hand—ready for mobile devices, drones, laptops, tools, medical devices, even electric vehicles (EVs). No one has yet, or ever will, come even close, he explained.
For example, ChargEdge alone had understood and dismantled a key foundational flaw crippling all of WPT today—the "double resonator" approach, as Sanjaya demonstrated through Mathcad. Everyone seems to have somehow automatically assumed, a priori, that "tuning both the receiver and transmitter to the 'same' frequency, independently, yields maximum power transfer." He explained why that was counterproductive, merely a reflex action based on years of experience in radio-frequency systems, which bear no resemblance to closely coupled systems such as WPT. He traced its destructive path from the earliest patents in 2002, still assigned to Amway, parent of the Wireless Power Consortium (WPC) and the Qi standard, right up to the 2019 patents of HaloIPT, formerly Qualcomm, now Witricity, on which almost all modern EV charging attempts rely. Sanjaya revealed that the elimination of that elemental flaw was the earliest point of digression in their development, which has now been woven into the fabric of their vast patent portfolio, guarded by White and Case. The recovery path for all others, stands closed!
Sanjaya demonstrated that the above design approach, rather design regression, was further compounded by one of the largest companies on earth, in a knee-jerk attempt to mitigate electromagnetic interference (EMI) by shoehorning "fixed-frequency switching" into the world of WPT. That too was just a conditioned reflex, based on years of experience in conventional switching power conversion. It can lead to severe drops in efficiency when dealing with WPT systems, as Sanjaya demonstrated via SIMPLIS simulations. Surprisingly, it was formalized into the Qi standard recently, and in fact forms the basis of every single wireless EV charging attempt over the last decade.
When things didn't go as expected, rather than question their false premises, they took refuge behind self-fulfilling postulations such as: "inductive systems offer inherently poor user experience compared to resonant systems." Ignoring the obvious fact that everyone, without exception, "tunes" both receiver and transmitter to the "same resonant frequency,"however erroneous. Including Qi.
Sanjaya wondered aloud, what else was WPC planning to resonate further, to be able to rebrand their self-described "inductive" standard as a "resonant standard?"—a clear reference to their constant press releases since 2014 claiming the imminent release of a "resonant extension to Qi." Which was meant to have been AirPower!
In a lighter vein, Sanjaya likened the course of modern WPT to the "Nixon Principle:" If two wrongs don't make a right, try three! Maybe four?
The unfortunate sequence occurred because they all consistently failed to realize the rules of the game are very different when it comes to resonance. For example in WPT, there are conduction losses but no switching losses—if implemented correctly (as ChargEdge alone guarantees). And so, the old paradigm that "higher frequencies lead to higher switching losses and thus lower efficiency," is opposite of what really happens in WPT. He revealed that he had taken deliberate care over the years to reexamine, if not entirely unlearn, traditional power conversion, despite having written seven very well-known books on that very subject. For example, the Chinese translation of Sanjaya's best known book, Switching Power Supplies A-Z, Second Edition, from Elsevier, holds an unprecedented 99.6% approval rating after almost 3000 reviews on Dangdang. But it is his other book, Switching Power Supply Design and Optimization, Second Edition, which features most of the scaling laws and resonant converter design which he revealed that evening.
Sanjaya warned against investing in impressive lab experiments, especially from academicians, because they usually ignored regulatory boundaries imposed by UL and FCC, and may have no hope ever of commercial release. He indicated those boundaries too can have major design consequences, and gave the example of Intel and IDT (now Renesas) among those investing heavily in the now-defunct Rezence standard, promoted by Witricity for years. Somehow, at the highest levels, they assumed: "higher frequencies will lead to smaller components"—a truism of conventional power conversion, but one which needs to be re-examined against regulatory limits. For example, in reality Rezence used gigantic air-cored transmitter and receiver coils with distributed capacitance, far bigger than even those of Qi, despite operating at a frequency fifty times higher—locked in a flailing effort to comply with the extremely tight 0.2% frequency tolerance mandated by FCC at their chosen frequency. Ferrites and commercial capacitors were unusable, offering at best 5-10% tolerance. Intel eventually abandoned Rezence abruptly in 2016.
Most surprisingly perhaps to all present that evening, the ChargEdge receiver coil, was visibly "smaller than half a pinky,"yet delivered live, up to 80W. Indeed, it was 10 times smaller than a 5W Qi coil, delivering 16 times more power. Yes, and it operated at around 100kHz, about the same as Qi, Sanjaya clarified in response to an audience member. On request, a volunteer in the audience further confirmed that the tiny coil was "barely even warm" to the touch, just moments after the massive 80W it had delivered.
The same transmitter offers "better than backward Qi compatibility," such as a proprietary Dual-Qi mode suitable for cars, and also charging of wearables from their sides, rather than through a hot back plate pressed against an inevitably degrading battery, as exemplified by Qi. Plus, the "world's smartest FOD" (foreign object detection), requiring no Rx-Tx communication whatsoever.
He pointed out that Qi's interoperability failure rate had skyrocketed above 60%, as reported by their own test labs, largely due to its flawed "15W FOD." In effect "there is now just a one out of three chance that your $1000 'fast charging' phone will even start charging consistently on an expensive 'fast Qi transmitter.'" And of course, you will get only a handful of watts anyway for all your trouble! Besides a case of alignment anxiety.
Because Qi has never delivered more than 3.5-4.7 Watts in practice, to any phone out there, as confirmed by several independent test labs. The reason being the hot receiver coil pressed against the battery, whose protection circuitry rushes in to try and limit the temperature rise and inevitable degradation, by reducing the accepted power.
Sanjaya unveiled yet another scaling law, the "V^2/R law," to explain that the live demo he had just presented—80W wirelessly transferred from a transmitter getting power from a small portable power bank providing just 12V, as confirmed by an audience member, implied that ChargEdge could at this very moment scale well beyond 50kW at automotive voltage levels, whereas all the others would get stuck at about 6-10 kW, excluding manual tuning on a case-by-case basis. As vindication, Sanjaya pointed to the only wireless EV charging "release" out there after a decade and millions spent, the BMW530E, based on Witricity technology, managing 3.2kW across 8 centimeters. More disturbingly, it releases almost half a kilowatt of heat within the narrow confines of the garage, while the owner sleeps unsuspectingly. In contrast, ChargEdge has extremely low receiver-side resistance and is also inherently alignment-friendly, Sanjaya pointed out.
ChargEdge went well beyond the industry, not only in its ability to locate, track and optimize around the actual location of the resonant peak, but in terms of preferred magnetic structures. Their transmitter structure is inherently zero-EMI, and the tiny receiver coil is virtually geometry agnostic, mountable on the edges of a phone for example, not pressed against the battery, halving its life as Qi does. The tiny coil can be planted within a wide variety of devices and even on the underside of EVs.
Phones can once again be predominantly metal too!
A scintillating Q&A session ensued, extending into the parking lot of the former National Semiconductor headquarters, where in fact Sanjaya had worked for five years, during which time he had introduced a well-known topology-unifying industry concept known as "current ripple ratio"—via two seminal application notes, AN-1197 and AN-1246, now belonging to Texas Instruments. With that, the evening came to a memorable end.
For example, ChargEdge alone had understood and dismantled a key foundational flaw crippling all of WPT today—the "double resonator" approach, as Sanjaya demonstrated through Mathcad. Everyone seems to have somehow automatically assumed, a priori, that "tuning both the receiver and transmitter to the 'same' frequency, independently, yields maximum power transfer." He explained why that was counterproductive, merely a reflex action based on years of experience in radio-frequency systems, which bear no resemblance to closely coupled systems such as WPT. He traced its destructive path from the earliest patents in 2002, still assigned to Amway, parent of the Wireless Power Consortium (WPC) and the Qi standard, right up to the 2019 patents of HaloIPT, formerly Qualcomm, now Witricity, on which almost all modern EV charging attempts rely. Sanjaya revealed that the elimination of that elemental flaw was the earliest point of digression in their development, which has now been woven into the fabric of their vast patent portfolio, guarded by White and Case. The recovery path for all others, stands closed!
Sanjaya demonstrated that the above design approach, rather design regression, was further compounded by one of the largest companies on earth, in a knee-jerk attempt to mitigate electromagnetic interference (EMI) by shoehorning "fixed-frequency switching" into the world of WPT. That too was just a conditioned reflex, based on years of experience in conventional switching power conversion. It can lead to severe drops in efficiency when dealing with WPT systems, as Sanjaya demonstrated via SIMPLIS simulations. Surprisingly, it was formalized into the Qi standard recently, and in fact forms the basis of every single wireless EV charging attempt over the last decade.
When things didn't go as expected, rather than question their false premises, they took refuge behind self-fulfilling postulations such as: "inductive systems offer inherently poor user experience compared to resonant systems." Ignoring the obvious fact that everyone, without exception, "tunes" both receiver and transmitter to the "same resonant frequency,"however erroneous. Including Qi.
Sanjaya wondered aloud, what else was WPC planning to resonate further, to be able to rebrand their self-described "inductive" standard as a "resonant standard?"—a clear reference to their constant press releases since 2014 claiming the imminent release of a "resonant extension to Qi." Which was meant to have been AirPower!
In a lighter vein, Sanjaya likened the course of modern WPT to the "Nixon Principle:" If two wrongs don't make a right, try three! Maybe four?
The unfortunate sequence occurred because they all consistently failed to realize the rules of the game are very different when it comes to resonance. For example in WPT, there are conduction losses but no switching losses—if implemented correctly (as ChargEdge alone guarantees). And so, the old paradigm that "higher frequencies lead to higher switching losses and thus lower efficiency," is opposite of what really happens in WPT. He revealed that he had taken deliberate care over the years to reexamine, if not entirely unlearn, traditional power conversion, despite having written seven very well-known books on that very subject. For example, the Chinese translation of Sanjaya's best known book, Switching Power Supplies A-Z, Second Edition, from Elsevier, holds an unprecedented 99.6% approval rating after almost 3000 reviews on Dangdang. But it is his other book, Switching Power Supply Design and Optimization, Second Edition, which features most of the scaling laws and resonant converter design which he revealed that evening.
Sanjaya warned against investing in impressive lab experiments, especially from academicians, because they usually ignored regulatory boundaries imposed by UL and FCC, and may have no hope ever of commercial release. He indicated those boundaries too can have major design consequences, and gave the example of Intel and IDT (now Renesas) among those investing heavily in the now-defunct Rezence standard, promoted by Witricity for years. Somehow, at the highest levels, they assumed: "higher frequencies will lead to smaller components"—a truism of conventional power conversion, but one which needs to be re-examined against regulatory limits. For example, in reality Rezence used gigantic air-cored transmitter and receiver coils with distributed capacitance, far bigger than even those of Qi, despite operating at a frequency fifty times higher—locked in a flailing effort to comply with the extremely tight 0.2% frequency tolerance mandated by FCC at their chosen frequency. Ferrites and commercial capacitors were unusable, offering at best 5-10% tolerance. Intel eventually abandoned Rezence abruptly in 2016.
Most surprisingly perhaps to all present that evening, the ChargEdge receiver coil, was visibly "smaller than half a pinky,"yet delivered live, up to 80W. Indeed, it was 10 times smaller than a 5W Qi coil, delivering 16 times more power. Yes, and it operated at around 100kHz, about the same as Qi, Sanjaya clarified in response to an audience member. On request, a volunteer in the audience further confirmed that the tiny coil was "barely even warm" to the touch, just moments after the massive 80W it had delivered.
The same transmitter offers "better than backward Qi compatibility," such as a proprietary Dual-Qi mode suitable for cars, and also charging of wearables from their sides, rather than through a hot back plate pressed against an inevitably degrading battery, as exemplified by Qi. Plus, the "world's smartest FOD" (foreign object detection), requiring no Rx-Tx communication whatsoever.
He pointed out that Qi's interoperability failure rate had skyrocketed above 60%, as reported by their own test labs, largely due to its flawed "15W FOD." In effect "there is now just a one out of three chance that your $1000 'fast charging' phone will even start charging consistently on an expensive 'fast Qi transmitter.'" And of course, you will get only a handful of watts anyway for all your trouble! Besides a case of alignment anxiety.
Because Qi has never delivered more than 3.5-4.7 Watts in practice, to any phone out there, as confirmed by several independent test labs. The reason being the hot receiver coil pressed against the battery, whose protection circuitry rushes in to try and limit the temperature rise and inevitable degradation, by reducing the accepted power.
Sanjaya unveiled yet another scaling law, the "V^2/R law," to explain that the live demo he had just presented—80W wirelessly transferred from a transmitter getting power from a small portable power bank providing just 12V, as confirmed by an audience member, implied that ChargEdge could at this very moment scale well beyond 50kW at automotive voltage levels, whereas all the others would get stuck at about 6-10 kW, excluding manual tuning on a case-by-case basis. As vindication, Sanjaya pointed to the only wireless EV charging "release" out there after a decade and millions spent, the BMW530E, based on Witricity technology, managing 3.2kW across 8 centimeters. More disturbingly, it releases almost half a kilowatt of heat within the narrow confines of the garage, while the owner sleeps unsuspectingly. In contrast, ChargEdge has extremely low receiver-side resistance and is also inherently alignment-friendly, Sanjaya pointed out.
ChargEdge went well beyond the industry, not only in its ability to locate, track and optimize around the actual location of the resonant peak, but in terms of preferred magnetic structures. Their transmitter structure is inherently zero-EMI, and the tiny receiver coil is virtually geometry agnostic, mountable on the edges of a phone for example, not pressed against the battery, halving its life as Qi does. The tiny coil can be planted within a wide variety of devices and even on the underside of EVs.
Phones can once again be predominantly metal too!
A scintillating Q&A session ensued, extending into the parking lot of the former National Semiconductor headquarters, where in fact Sanjaya had worked for five years, during which time he had introduced a well-known topology-unifying industry concept known as "current ripple ratio"—via two seminal application notes, AN-1197 and AN-1246, now belonging to Texas Instruments. With that, the evening came to a memorable end.
