Untangling in-vehicle wireless power design challenges

Article By : Chris T. Burket

Car model implementation numbers, with wireless charging available, have grown from 40+ in early 2016, to 100+ models (currently).

For those of us who, in our daily commute, spend an inordinate amount of time in stop-and-go traffic (or completely parked on those things called Interstate Highways), know that this idle time can be used for many other purposes, and that the automobile has always been a key market for technologies that assist us in being able to do some of those other “tasks,” such as phone conversations, texting and e-mailing, online shopping and surfing, movie downloading and video streaming, game playing, and more. If you look left or right while in traffic, you can see that a majority of these tasks evolve around the use of the handset.

To supplement this high usage, one of those newer implemented technologies is in-vehicle wireless charging functionality built into the center console area or some other easily accessible location. The purpose? To remove all those plug-in cables and to have the handset sitting in a known location while being charged.

Background

Within the past three years, the wireless power technology “wars” were resolved with “Qi” or Wireless Power Consortium (WPC) being the winner, and now the low power de-facto standard. This was further validated with the implementation of the Qi technology by all leading worldwide handset makers. Prior to this, automakers did implement wireless charging within their vehicles, but there was always concern as to whether a sale would be lost solely because the buyer’s phone was not compatible with the embedded charging technology.

Car model implementation numbers, with wireless charging available, have grown from 40+ in early 2016, to 100+ models (currently), which equates to over 12 million (2.4M OEM installed, 9.7M after market) vehicles having Qi based in-vehicle systems installed in 2018 alone. The majority of those systems were compliant with the Qi Basic Power Profile (BPP) and for 5W (watts). The new direction is for faster charging and higher power. A majority of new designs are targeting compliance with the Qi Extended Power Profile (EPP) or 15W capability. This added convenience of being able to charge faster comes with additional technical hurdles that must be overcome. The three main issues being EMI compliance, efficiency and thermal limitations.

The 15W System

Within the WPC standard, there are sub-categories (e.g. MP-A8, MP-A9, MP-A13) that specify various aspects of the wireless power system and the configuration of the transmit (Tx) coil that is placed within the center console area. Done for interoperability purposes, the standard defines: input DC voltage, Tx coil size and shape, electrical parameters, frequency control (fixed versus variable), power level, and power control (voltage/frequency/phase/duty cycle). The input voltage, using the vehicle’s main battery, is typically 12V into the transmitter’s circuitry and thus has an elevated voltage, creating a stronger electrical field (E) than the 5V input voltage associated with many desktop wireless chargers. Due to the resonance mode of operation within the system, the actual voltages on the coils (resonators/antennas) can be around 100V, meaning the radiated noise will be stronger for the 15W system than the 5W system.

EMI Issues and Solutions

On newer vehicles, there are numerous RF systems, all needing to co-exist to ensure that what they are doing does not affect anything else. Some of these being: AM/FM radio, GPS, ADAS systems, multiple cellular bands, Blue Tooth, WiFi, asset tracking, short wave radios, key fobs, police scanners, telematics, etc., and maybe, even a few CB radios for all those 10-4 buddies out there.

A few of these RF systems operate within the 87-205 KHz (can be up to 300 KHz) fundamental frequency range of the Qi EPP wireless power system and/or through the low harmonics. AM radio, 525 KHz to 1705 KHz (in the Americas), is required to be EMI free as it is used as part of the Emergency Broadcast System. New remote keyless entry systems (RKE) operate at 125 KHz as do some Tire Pressure Monitoring Systems (TPMS) which use this frequency to drive the initiator L-C coil circuit.

Automotive applications have very stringent requirements for EMI. CISPR 25 (Comité International Spécial des Perturbations Radioélectriques) is a non-regulatory engineering automotive standard that sets conducted and radiated emission limits that must be met for the protection of other on-board receivers. It defines these limits over a frequency range of 150 kHz to 2500 MHz which could possibly be conducted by other vehicle mounted antennas.

Within CISPR 25, there are classes that define the level of permitted conducted and radiated noise emissions limits, with radiated noise being the real concern. The Class emission [radiated] limits versus bands are given in Table 1 for Peak, Quasi-peak and Average measured voltage up through the FM radio band.

Table 1: CISPR 25 Radiated Limits by Class
Table 1: CISPR 25 Radiated Limits by Class

With the increased Qi EPP power levels, meeting Class 4 has been a challenge, and no Class 5 system is available in the market yet. For in-vehicle wireless charging, the AM frequencies up to 1.8 MHz are the most sensitive but certification testing does go up past 1 GHz. Actual CISPR 25 Class 5 measurement data is provided in Figure 1.

CISPR 25 Class 5 Initial Testing 100 KHz to 30 MHz
Figure 1: CISPR 25 Class 5 Initial Testing 100 KHz to 30 MHz

From the plot, this design does not fully pass Class 5 certification, though it does meet Class 4 requirements. EMI noise mitigation starts with the system’s electrical design and the following sections address some key areas used in the design to meet CISPR 25 requirements.

The first area of mitigating EMI noise is implementing a fixed frequency system. Within the Qi standard, there are means to allow for variable frequency to better “tune” the two sides for improved performance. However, to meet the stringent EMI noise levels associated with in-vehicle power systems, a changing frequency would make complying with these even more problematic. Also, European automakers have restrictions above 145 KHz so the fixed operating frequency of current solutions is set at around 127 KHz.

The next technique is to remove square wave currents though the Tx coil and have these currents as close to sinusoidal as possible. This approach reduces the noise “spikes” that could be otherwise generated. This can be achieved by using an inductor as this passive device smooths out the square wave current created by the turning ON/OFF of the switches (MOSFETs), and helps to ensure that the switching scheme is “clean” and noise free.

Further EMI suppression can be realized with the addition of a common mode filter (CMF) placed on the power lines in series with the Tx coil windings. Currents through the coils are 100% alternating current (AC) and have no direct current (DC) content like many power supplies that involve DC current and some allowable ripple current. The coil’s current can be thought of as being 100% ripple current. Therefore, selection of the ferrite material used for this CMF is important and AC core loss must be at an absolute minimum at the 127 KHz fixed frequency.

Another EMI noise suppression technique is to add EMI noise suppression magnetic sheets to absorb operating frequency, harmonics and spurious noise generation that may be transmitted out of the backside of the main Tx shield. The magnetic sheets remove EMI noise via two methods. First, the permeability (µ’) of these materials enables these shields to contain [absorb] the EMI noise magnetic flux (φ) and keep it from being radiated. Next, the resistive properties (µ”) of these shields create a resistive path for the unwanted frequencies’ flux field and attenuate the EMI noise and remove it from the environment in the form of heat. This relationship is given in Equation 1.

µ = µ’ – jµ”                          [Eq.   1]

For EMI suppression applications, higher µ’ yields better shielding performance through magnetic flux containment and higher µ” yields better noise suppression through material core losses. Having too high of a µ’ value can decrease performance. Due to a phenomenon called magnetic coupling (K), having an additional magnetic sheet can shift the inductance value of the Tx coil and de-tune the circuit through mutual coupling (M or Lm) and move it away from the desired fixed frequency.

Lastly, if EMI suppression sheets do cause fixed frequency problems, there are non-magnetic materials that also suppress EMI noise. The challenge is to obtain a material that can absorb some level of noise energy, yet is not too metalized as to simply reflect the EMI noise rather than removing it and not suppress the desired H field. Silver alloy-based films with low surface resistances (~4 ohms/square) have been used and demonstrate improved EMI noise suppression up to 1 MHz, (I think “and” needs to be included) dampening of problematic harmonics. These non-magnetic sheets, placed on the top side of the windings, tend to better suppress voltage/E field-based harmonics rather than current/H field-based harmonics.

The Tx coil comes with its own magnetic shield which contains magnetic flux generated by the sinusoidal electrical current going through the winding. For the fundamental operating frequency (127 KHz), the shielding material is selected to have higher µ’ and very low µ” as to not attenuate the desired magnetic flux field. This shield contains the wanted magnetic flux at the operating frequency for improved performance and some of the harmonic flux, thus becoming part of the of the overall EMI compliance solution.

Efficiency – Wireless Power System Factors

With no direct electrical connection between charger (transmit side or Tx) and the “to-be-charged” receive (Rx) device, the energy is transferred between the two sides via an H field created by the electrical current flowing through the Tx coil. The Rx coil captures a portion of this H field and converts it to an electrical current through the Rx winding.

The mechanism for this process is magnetic coupling, and is impacted by alignment between the two coils (X, Y directions), the separation distance (Z gap) and the orientation (parallel). In a center console automotive application, the orientation issue is controlled by the flat console area surface.

Alignment is addressed currently by a 3 distinct windings Tx coil pattern and, with some built-in control intelligence, the Qi system determines which winding is best aligned. This 3 windings coil provides some level of position freedom, but only in one axis. An example of an MP-A9 coil is shown in Figure 2.

Standard WPC MP-A9 Tx Coil
Figure 2: Standard WPC MP-A9 Tx Coil

This 3 windings coil example shown is not a system requirement, but has been the “norm” to date. However, there are current efforts to have a 2 windings coil configuration for smaller vehicles to reduce both size and cost. The trade-off will be in the alignment of the Rx coil with one of the two Tx windings to ensure efficiency is not impacted by reduced magnetic coupling. The typical minimum system efficiency requirement is 70% from the automakers.

The Z direction gap is more of a challenge due to the original Qi standard specifying the maximum distance between the two sides as <5 mm, and is for the magnetic shield-shield distance, not for coil winding-winding. The shields are used for: 1) shielding the H field from objects behind the coils, 2) shaping/directing/encapsulating the H field, 3) helping to set the inductance values, and 4) providing the mechanism for magnetic coupling, a function of the physical distance between the two magnetic sheets. Therefore, for an automotive application with a winding thickness of up to 1.0 mm on the Rx coil, a 1.0 mm phone back cover thickness, a phone protection case up to 3.0 mm or more, a center console thickness of 2.0 mm, and a Tx winding structure thickness of 2.5 mm, all means the overall shield-shield distance far exceeds that 5.0 mm limit. The actual Z gap distance is more along the lines of 9-10 mm, which assumes no gap between the phone case and the center console. As the coupling factor decreases with Z gap distance, pressure is put unto the Tx side to compensate for the lower coupling by requiring more Tx input current to maintain the power required on the Rx side, since the electrical load does not change. Requiring more input current to maintain the same output power is another way of saying decreased efficiency. This is shown in Figure 3.

Efficiency Vs. Z Gap using WPC A11 Tx Coil, 5V Rx Output
Figure 3: Efficiency Vs. Z Gap using WPC A11 Tx Coil, 5V Rx Output

This test data was done using a 5W Qi A11 Tx coil, two Rx coil sizes and with/without having a battery located behind the Rx coil. The efficiency rolls off as the Z gap increases between the coils. The Qi system also operates with in-band communication and when coupling (K) is low, there is a chance communication may cease and stop power transmission. This was the case for the orange triangle curve (lowest) and therefore all testing was terminated at 11 mm. For Qi EPP systems, the higher currents will help with coupling, but it is important to understand what the real-world Z gap is.

Higher coil winding currents also create higher wire losses. There is both a coil’s DC resistance (DCR) value and AC resistance (ACR or Rac) value, with wire loss being related to ACR and is shown in Equation 2.

PLOSS = I2 * RAC                                 [Eq.n2]

where:

I = AC current though the Tx coil winding

RAC = resistance at some given frequency

Higher current systems generate more magnetic flux (φ) and higher core losses within the Tx coil magnetic shield.  Typical curves, for various magnetic materials, are shown in Figure 4 for core loss (Pcv) versus magnetic flux density (B), where nickel-zinc (Ni-Zn) and manganese-zinc (Mn-Zn) are types of ferrite.

Magnetic Shield Core Loss vs. Magnetic Flux Density power
Figure 4: Magnetic Shield Core Loss vs. Magnetic Flux Density

Core loss is a function of the magnetic field flux density (B) within the core. Flux density is related to the magnetic field (H) and is supplied by magnetic core suppliers in the material’s B-H curve. The relationship between increased current and H field is given in Equation 3.

H ∝ N x I                            [Eq. 3]

where,

N – the number of winding pattern turns on the coil

I – current through the winding (A)

In wireless charging systems, the alignment and Z gap parameters cannot be 100% controlled by either the automaker, nor the Tx charger system maker. How and where the user places their handset into the console area, what type of protective case is used, internal Rx coil size and shape, and whether the handset shifts during acceleration and braking, all will impact efficiency.

Efficiency – Tx Power Topology Factors

A key efficiency improving technique is to use a Push-Pull converter drive scheme. The Push-Pull converter supplies the Tx coil current by a set of switches in a synchronized timing scheme.

The switches are alternately switched ON and OFF, thereby cycling the direction of the current through the coil during both halves of the switching cycle, unlike Buck, Boost, and other topologies that rely on stored energy within passive devices to supply current during the switch’s OFF period. Also implemented with Push-Pulls, is a short “dead” time with no current, ensuring both switches are not ON (sourcing current) at the same time, which would cause damage to the supply. Overall, Push–Pull converters have steadier input current over other power topologies, generate less EMI noise, and are more efficient in higher power applications.

A technique also used is called Zero Voltage Switching (ZVS) or “soft” switching. In order to reduce losses during the turning ON/OFF of the switches (MOSFETs), the system ensures that prior to the switching process, there is no voltage across the switch. This eliminates the possibility of having current flowing through the switch with an applied voltage. Having ZVS reduces switching losses and substantially improves efficiency. Therefore, timing control is a key requirement.

Another advantage of ZVS is that it helps to reduce harmonics. Reducing the harmonics assists in complying with the CISPR 25 requirements discussed above. This is included here, as it plays a key part in efficiency. In an inductive wireless power system, the Tx and Rx side circuits’ L-C (inductive-capacitive) networks are tuned to a specific frequency. Energy in harmonics must be minimized as it is wasted energy, since the tuned Rx side will not rectify any energy outside of its tuned frequency range.

Thermal Issues

Losses in wireless power systems come from circuit components on both the Tx and Rx sides, wire and core losses on both coils and finally, coupling losses to bridge the air gap between the two coils. Within the auto-maker’s control are the Tx side wire and core losses, and the circuit component losses, including the PCB. All these losses lead to increased temperature rise. Automakers have very stringent temperature rise limitations that are usually at +10oC rise over ambient temperature.

Magnetic core losses are dependent on the material’s characteristics. The key parameters are which type of material is used, i.e., ferrite or powdered iron, the thickness of the shield, the operating frequency (which creates inner-material hysteresis and Eddy current or “spin” losses), and the magnetic field flux density. Temperature can have a big impact, but limitations set by the automakers have ensured that the temperature does not remotely approach a core temperature to have any real effect due to touch temperature safety issues.

To decrease Tx wire losses, a multi-strand litz wire is used to reduce AC losses as the frequency increases, due to a phenomenon called “skin effect”. Simply stated, as the frequency increases, more current flows closer (therefore higher current density) to the wire’s outer surface and utilizes less of the wire’s cross-sectional area. This creates increased resistance, which increases wire loss and temperature rise. Litz wire creates much more of an overall wire surface area and helps reduce AC resistance over standard single strand wire. As the power and current requirement levels continue to go up, the need to use higher stand litz wire increases. Balancing performance, thermal issues, wire diameter and coil size then become key aspects of in-vehicle systems.

Heat sinking is used as a way of removing Tx side heat. However, the automakers resist any extra heat sinks to be added as it increases weight, so the push is prioritizing efficiency.

This article originally published on our sister site, Power Electronics News

Virtual Event - PowerUP Asia 2024 is coming (May 21-23, 2024)

Power Semiconductor Innovations Toward Green Goals, Decarbonization and Sustainability

Day 1: GaN and SiC Semiconductors

Day 2: Power Semiconductors in Low- and High-Power Applications

Day 3: Power Semiconductor Packaging Technologies and Renewable Energy

Register to watch 30+ conference speeches and visit booths, download technical whitepapers.

Leave a comment