RoadTest: Wireless Power Design Kit - Medium Power
Author: shabaz
Creation date:
Evaluation Type: Development Boards & Tools
Did you receive all parts the manufacturer stated would be included in the package?: True
What other parts do you consider comparable to this product?: null
What were the biggest problems encountered?: No issues encountered, but as a personal preference it would have been nice if the solution was designed around more popular micro-controllers
Detailed Review:
Wireless power enables all sorts of devices to be powered, or charged, by placing them on a charging pad. It has some distinct advantages which will be explored further below.
Early experiences of it were not that great, until the products and the standards matured. Older mobile phones needed a bulky attachment to enable wireless charging. Recent smartphones have a high power demand and the older wireless power standard would only offer 5W; this meant that it could take many hours to charge a phone.
I was therefore interested to try out the new Wurth Medium Power design kitWurth Medium Power design kit that offers 15W, which is a far more interesting amount of usable power for many applications. The kit contains the wireless power transmitter and the power receiver. The photo here shows the receiver sitting on top of the transmitter board, illuminating a white LED load which is also supplied.
The term ‘wireless power’ can sound a bit vague, so in a nutshell the type of wireless power technology that is to be discussed here, we can consider to be a system of transferring energy wirelessly in a controlled, practical and interoperable manner for providing sufficient energy to power or charge some typical (often portable and handheld) modern electronic products.
The kit comes in a plastic box with shaped foam inside; many Wurth kits use this method, it is great : ) Often they come with handy additions like plastic tweezers too. This is my third Wurth kit, I have purchased a 5W Wireless Power kit in the past, and a power supply kit.
The kit also contained a mains power supply with mains plug attachments for different regions.
The photo here shows the new 15W design kit, and the older 5W kit.
There are several standards-based methods of wireless power, some competing and some intended for different use-cases. The benefits mentioned here are fairly common to all of the methods, but I’ll predominantly focus on the Qi (pronounced ‘chee’) standard, which is a popular way of implementing wireless power for smartphones these days. Some other wireless power methods not discussed here include AirFuel, which includes technology intended for charging earphones placed inside a case.
The Qi standard uses thin coils to implement a power transmitter and receiver system, and smartphones such as the iPhone 8, 8 Plus and iPhone X, and Samsung Galaxy S6, 7 and 8 have the receiver coil built-in.
The system uses inductive power transfer which doesn’t interfere with mobile phone reception. It is also very efficient; an end-to-end system efficiency of 70% is not unusual, and it is possible to exceed 90%. Qi originally supported just 5W capability but the newer version of the standard (version 1.2.x) supports Extended Power Profile (EPP) which offers up to 15W power capability.
The inductive system uses flat wound coils and often ferrite material. These are well-known technologies and it makes the solution cost-effective. In theory in future devices may not have any charging connector; this would reduce costs because connectors are expensive and sometimes difficult to integrate into an enclosure, especially if dust and water sealing is needed.
The Qi system in particular allows for several implementations; a basic version may require the receiving power device to be positioned accurately, but more advanced versions can implement a larger mat where the device can be positioned anywhere on it.
The system uses data communication from the power receiver to the power transmitter and thus the system is safe; it won’t provide (significant) power to any receiving coil placed on the mat, unless it communicates back in a defined manner.
While smartphone charging is a popular use-case, it is possible to imagine Qi technology being useful for powering equipment in more unusual environments since no connector is needed. It is also great in vehicles, where it would be awkward for someone to plug in a phone while it is moving or to repeatedly un-dock a connector during package deliveries. There could be industrial, food processing and other use-cases where instruments need powering but sparking or contamination or dirt on metal connectors could be an issue.
The system uses primary and secondary coils, much like a transformer. The primary coil acts as the power transmitter, and the secondary coil is the receiver. Alternating current in the primary coil causes changing magnetic flux, and this causes a current to be created in the secondary coil. The system operates at around 100kHz. The coils are paired with a capacitor in the circuit, and this forms a ‘tank circuit’ which will resonate at a particular frequency. The transmit and receive circuits are both tuned to the same frequency.
The ratio of coil sizes and distance between them and other factors can affect the level of power that can be transferred. Rather than tightly defining the transmitter and the receiver, instead the Qi standard has requirements for the transmitter side, and this then enables manufacturers to implement the receiver more flexibly, to interoperate with such ‘reference’ transmitters.
The general guidelines for the system are that it operates most efficiently when the two coils are of similar size, and the gap between them is small and they are as concentric as possible when viewed from above.
There could be scenarios where a larger pad is desired. Solutions for this can involve having multiple smaller coils within the pad. It is still efficient, because power is only enabled in any significant amount to a few coils, based on detection of where the device is on the pad.
When the coils are of similar size, the gap between them can easily reach 20% of the coil diameter with little loss of efficiency. This translates to up to about 7mm of gap possible with the Wurth coils used in the design kit. In addition, for such a size of coil, it should be feasible to have around 5-10mm of displacement between the centres of the coils with little loss in efficiency. In other words, the charging area would have twice this amount, i.e. 10-20mm of overall tolerance when the user placed the device to be charged onto the charging pad. This type of single-coil transmitter design is suitable where the charging pad is small and it is easy for the user to centre the device manually to within this tolerance. This is not bad, and easy for a user to do visually. In some situations, the pad could be moulded to fit around the device to be powered or charged, and that would ensure that the coils are well centered. An example would be a cradle for a mobile phone in a car. Another example could be to use magnets to snap the pad and the device to be charged into position. As mentioned, an overall tolerance of 10-20mm is fine though.
There are four phases that occur in order to successfully achieve wireless power transfer with the Qi system. There are a couple of additional phases for power requirements beyond 5W.
During this phase, the transmit coil pulses power (an AC signal at the resonant frequency) to the transmit (i.e. primary) coil. The power output is at a low level.
The aim of this phase is to detect when a user places a receiving circuit on top. When that occurs, power is induced in the receiving coil, and causes the receiving tank circuit to resonate too. So, when the transmit coil is powered off, the receiving circuit continues to resonate briefly. This can be detected by the transmit side while it is not transmitting.
Once the transmit side has detected this signal, it knows that there is some resonant circuit that could require power transfer, and so the system can move on to the next phase.
During this phase, the transmitter (i.e primary coil) side sends enough AC power (continuously, not briefly like the first phase) to allow the secondary (i.e. power receiver) side to power up and send back digital information in the form of a signal strength (SS) packet.
The information is sent by applying/removing a load. The primary side can detect when a load is applied by the secondary side, because the AC voltage drops (and the current increases) in the primary side. The information sent specifies the received signal level. The primary side can use this information in later phases to adjust the power output accordingly.
Up to some 27 bytes of information can be sent in packets during this phase by the secondary side. The information includes version and identification information and the maximum power that is needed. The primary side uses this information to see if it is compatible with the secondary side. If it is, then the next phase is entered.
This phase is only used in the extended power profile (EPP) mode. There is bidirectional digital communication during this phase to determine which power capabilities the receiver requires and acceptance of the power transmitter to supply it.
This phase is only used in the EPP mode, and is intended to make the foreign object detection (FOD) more accurate, by sending received power information for a light load and a heavy load, so that the transmitter side can more accurately determine if a foreign object is present. The foreign objects that are of concern are things like coins, which someone may place on the charging surface; it would be bad if they heated up and caused a danger!
During this phase, the primary side increases the transmitted power, and the secondary side sends back an error signal to allow the primary side to adjust the power level accordingly.
During this phase the secondary side has sufficient power to allow the device to run, or to charge a battery as an example. In the case of the demo kit, during this phase, the LED load is expected to be lighted.
The design kit is in two main parts; the transmitter portion which acts as a charging pad, and the receiver portion which is the device to be powered or charged.
The transmitter portion can be considered to be a reference design for a charging pad.
As can be seen there is quite a lot of circuitry, not everything is integrated into a single IC. Rohm’s BD57020MWV integrated circuit performs the Qi operations and is controlled and monitored by a microcontroller via I2C. The BD57020MWV has close to 50 registers that can be read or written.
The external circuitry includes a H-bridge to drive the coil, but also in particular there is circuitry for the foreign object detection feature. This is achieved in two ways. One way is to measure the current and voltage being supplied to the coil. An external current sensor (a TI part) is used for this. There is also a ‘Q factor’ measurement that is done, and it uses an op-amp and discrete circuit. The Q factor is the ratio between the signal amplitude across the inductor during resonance, and the input signal amplitude. It indicates how sharp the amplitude difference is, when approaching the resonant frequency (which is 100kHz for the inductor/capacitor combination).
The Q-factor based FOD functions by measuring the amplitude of the resonance across the coil to compute a Q-factor value when there is no metal near the coil, and then when a receiver is nearby it will measure the amplitude of the resonance again. If there is a significant drop in Q, then that means there is a foreign object.
The Wurth coilWurth coil consists of an impressively large slab of ferrite with a dual layer Litz wire winding. It is secured to the PCB with an adhesive foam pad. Total thickness including the pad is 6mm. The enclosure of the charging pad is simulated with a 2mm thick transparent sheet of Perspex mounted about 1mm off the surface of the coil. In other words, any device to be charged will be at least 3mm away from the primary coil. A real charging pad could have a reduced distance with a moulded enclosure.
The coil looks superb; very precisely wound for repeatable results, and very substantial Litz wire for a large surface area, for efficient current-carrying capability at the frequency of operation. The great thing about Wurth coils is that they always publish intensive measurements. Some manufacturers don’t. So, it is easy to see what Q-factor the coil possesses by checking the datasheet; it is an extremely impressive value of 180 at the relevant frequency (close to 100kHz). A high value will help to make the Q-factor based FOD function accurately.
In summary while the design is not as integrated as hoped, for anyone interested in creating wireless power products, the transmitter portion of the design kit makes an excellent reference transmitter design to test the products with, because all the relevant measurement points are either already accessible or can be made accessible (for example, it was easy to tack some wires onto the board in order to obtain the oscilloscope traces further below). There are also useful LED indications of key states such as low power or medium power transfer, and error conditions.
In terms of things I'd have liked to see different, personally I would have preferred to see a more popular microcontroller used, so that direct access and modification to the registers was easy too. The device used on the transmitter board is a Rohm 8-bit microcontrollerRohm 8-bit microcontroller. That's just me, others may be familiar with this device.
The receiver portion of the design kit has an interesting construction too. It consists of two PCBs. The lower PCB is the actual receiver, and the upper PCB contains a 64-LED load (described further below) for demonstrating the system.
The receiver board is fully integrated, with all active circuitry achieved in the single Rohm BD57015GWLBD57015GWL integrated circuit. The device uses an internal full-bridge rectifier based on MOSFETs instead of diodes (i.e. a synchronous rectifier) for very low loss, and an internal programmable low dropout (LDO) voltage regulator that can be set between 5V and 12V.
The receiver (i.e. secondary) coil is a Wurth charging coilWurth charging coil which occupies less footprint than the primary coil and is far thinner (overall thickness is 1.2mm including adhesive tape) and lighter too, ideal for mobile devices. It doesn’t need as high a Q, so it is wound differently, It uses a lower-cost pair of windings instead of Litz wire. I was very impressed at the small size of the secondary coil; it is not much bigger than coils intended for 5W power designs!
If different coils need to be tested, the plastic block that the receiver is mounted on can be unscrewed, and the two coil connections need unsoldering – this is very straightforward and well thought through, because the coil connections are brought out to the side of the PCB, so no holes to suck solder out of while trying many coils : )
It is important to note that with the Qi system, for transients, voltage at the input of the voltage regulator is ‘open loop’ and can vary depending on load. There is closed loop feedback to compensate for load, but it is relatively slow because it depends on digital information sent from the power receiver to the power transmitter, which then adjusts the power applied to the primary coil accordingly.
If the expected load has very fast-changing behaviour that the rectified output minus the LDO dropout voltage cannot handle, then an additional voltage regulator or DC-DC converter may be needed to handle it. Ordinarily this would not be an issue where the target load is a battery to be charged for instance, or an LED lamp where it doesn’t matter if there is a slight drop in voltage for a fraction of a second when the lamp is switched on.
It is a really nice feature that the load board is entirely separate and can be replaced with custom hardware. The supplied load module as mentioned earlier is comprised of a large array of white LEDs, back-lighting a Wurth/Rohm logo on white plastic.
The circuit board has an LED driver chip, and also a microcontroller that connects to the receiver board using I2C. The BD57015GWL Wireless Power Receiver IC is actually completely stand-alone and doesn’t require I2C communication if a basic 5W system is required. For higher power, a microcontroller is needed to configure for the higher current. As supplied, the load is configured for 10W but it can be reprogrammed to supply 15W. It requires new firmware to be uploaded, or (easier) just swap out the load PCB with your own desired hardware, and use any microcontroller of your choice to control the receiver board via I2C.
The supplied load has a total of 64 white LEDs connected to a BD6142AMUVBD6142AMUV integrated circuit which is intended for applications such as LCD backlighting.
It was interesting to examine the generated output from the wireless power transmitter, right at the coil, for some of the Qi phases.
When there is no receiver board nearby, the transmitter is in the Start Phase. This means it is actively seeking out if any receiver may be nearby. A frequent 100kHz resonant signal could be observed:
It is likely that this is the signal used to measure and record the Q of the coil when there is nothing else nearby, to be used for foreign object detection (see further below for an example of this).
As well as this frequent burst, there is a brief 68msec power transmission every 1.3 seconds; the frequency was about 175kHz:
If there is a load detected (for instance by picking up resonance from the receiver coil after the 68msec burst, then the next phase is entered.
It would be awkward to capture all the other subsequent phases (and I didn’t want to decode the entire Qi specification this way, since it is available to download for free) but from a load perspective the important phase is the final Power Transfer phase. To capture it, I placed the supplied receiver and load boards on top, and attached two oscilloscope probes to the two ends of the primary (i.e. transmit side) coil. The probe grounds were connected to 0V. The yellow trace in the oscilloscope views below shows the voltage at one end of the coil (it happens to be connected to the half of the H-bridge that is connected to the SW1 pin on the wireless power transmitter IC) and the blue trace is connected to the half of the H-bridge connected to the SW2 pin.
It can be seen that the signals are 180 degrees out of phase (although this isn’t always the case; the wireless power transmitter runs an algorithm that can result in the frequency and phase being different). The red trace is the result of the math calculation of the difference between the two, so that the actual voltage across the coil can be observed.
The oscilloscope was programmed to provide a frequency and a RMS measurement. The RMS measurement is indicated in red to be around 25.25V RMS.
If the load were to change, it would be expected to see the values and possibly the phase change too, such that the load meets demand (if you recall, the wireless power receiver sends back an error value digitally modulated on the load, so that the power transmitter side can optimise the power level; it is essentially a slow closed loop).
It appears the algorithm can find different phase and frequency combinations to achieve the desired output level; for example, I sometimes observed the signal across the primary coil to be as shown below:
Although the RMS value is different, this is likely partially due to precise load positioning having an effect on the amount of received power; I’m considering drilling holes in a few locations on the Perspex cover on the transmitter board, and adding small pegs on the wireless receiver, to explore this more accurately.
Increasing the gap between the two coils shows how the amount of transmit power dynamically changes as a result of the closed loop due to the digital information sent from the wireless receiver. I had a piece of 3mm plastic to use as a shim, but 3mm was found to be too large a gap for the system to initiate the power transfer, so I place the load onto the transmitter without the shim, and then when the Power Transfer stage was reached, I slid the plastic in carefully, and it maintained the power transfer : )
As expected, the voltage at the transmit coil increased, in this case to 30.8V RMS.
I have a Samsung S7 Edge phone which has medium power wireless charging capability, but unfortunately Samsung is using proprietary data; it is not compatible with the Qi standard except in the base low-power profile of 5W. Very disappointing : ( it means only ‘official’ wireless chargers for Samsung phones will provide fast wireless charging; all others will drop to 5W. However, I explored what would occur with the basic 5W profile. The screenshot below was captured with the phone placed on top of the transmitter. The key difference is that now only one of the half-bridges is being switched (the blue line doesn’t appear visible, but it is actually on the axis, at 0V). This is sufficient to transfer 5W of power. The interesting shape is due to how the S7 loads the system to achieve wireless charging.
I used the wireless power receiver from my old Wurth 5W kit with the new transmit board, and captured the following oscilloscope trace; again it can be seen that the 5W was supplied in a similar manner:
The way the Q factor plays a role in foreign object detection has been discussed, but I was curious to try to invoke it and see the waveforms. With a UK 5 pence coin placed in the centre of the detection area, I placed the receiver board on top. The power transmitter board detected the fault condition, and started rapidly flashing the FOD warning LED. It was now in a mode where it was waiting for the foreign object to be removed. On the oscilloscope, I could see that all that was occurring at the primary coil was frequent resonant oscillation tests due to impulses from the wireless power transmitter IC. The presence of the coin had sufficiently lowered Q, and it was visible on the oscilloscope trace (the white trace shows the greatly reduced resonance compared to the original yellow trace):
The Wurth Medium Power Wireless Power Design Kit has a valuable set of functionality for those looking to enable wireless power or charging capability in products. The on-board LEDs and test points allow for significant insight into how the wireless power receiver and load is behaving. The Wurth primary coil looks near-essential for high quality charger design; it will have good sensitivity for foreign object detection which makes the design safe.
The highlights were the fact that the receiver IC can function completely standalone (I2C is required for going beyond 5W though) and required few additional components, and the very light and thin secondary coil.
I also liked that the entire design kit disassembles really well so that different areas of the product being designed can be explored. One can start prototyping with the target load without needing to wait for a PCB containing the wireless power receiver to be manufacturered. The receive coil too is easily detachable, for testing other coils if desired.
Armed with this design kit, one would be very comfortable developing wireless power solutions. I'm actually really excited that with this kit, I've discovered that the Wurth charging coilWurth charging coil and BD57015GWL Wireless Power Receiver ICBD57015GWL Wireless Power Receiver IC together makes a very easy-to-use medium-power wireless power supply/charger solution for integrating into future designs.
Top Comments
Awesome as usual. I am saving this for future reference.
Hi Shabaz,
A very interesting read. I had a very rudimentary understanding and you have filled in a lot of questions for me. I had no idea that there was so much communication going on between the transmission…
Hi John,
Thanks!