The discussion focuses on designing a 5V DC to 5V AC, 5W wireless charging circuit operating at 275kHz, which is outside the typical 85kHz Qi standard. The recommended approach involves using resonant SMPS topologies such as an LLC half-bridge or a Hartley oscillator with a high-Q LC tank to minimize harmonic distortion and EMI, especially considering FCC regulations on AM band interference from harmonics. The transmitter coil should be larger than the receiver coil and placed as close as possible (ideally under 1 cm) to maximize coupling efficiency. Bluetooth control via an ESP32 is used for power management, and pre-made coils with ferrite backing plates are preferred for consistent inductance and efficiency.
Embedded systems engineers, power electronics designers, and hobbyists developing custom wireless charging solutions at non-standard frequencies who require guidance on resonant circuit design, EMI compliance, and coil coupling optimization.
What is your use case for this charger? It may not be good for phonesSterer said:
Over what distance? Can the charging coil be placed in direct external contact with the battery case? Or better yet, can it be designed so the battery power receiver part fits inside the charging coil, kind of like some electric toothbrushes are recharged?Sterer said:
As is the 2nd, 4th, and 5th. Typically the 3rd and 5th are higher in amplitude than the even harmonics. I believe I have this right. Feel free to correct me as I would put more faith in your knowledge of RF than pretty much anyone I know these days.berkeman said:
I would look into resonant SMPS designs. My first guess is a resonant LLC half-bridge topology, but I haven't looked into this for a few decades. There is a lot of information available and probably some control ICs. Look at TI.com and Analog.com (or google) for starters.Sterer said:
Agreed. I was just using the 3rd harmonic because it is usually the biggest depending on the signal generation circuit, and was an existence proof that there may be regulatory issues.Averagesupernova said:
Sterer said:
Sterer said:
berkeman said:
The problem is not creating a 5W 275kHz sinusoidal waveform with low harmonic content to drive a charging coil. The problem is optimizing your wireless charging setup to help you achieve your goal. If you give us so few details about your application, you are not likely to succeed. What exactly is your application and what are your design constraints?AlexB23 said:
Why?Sterer said:
I'm not sure what makes a relaxation oscillator "standard", but the operating frequency isn't hard to make whatever it is.Sterer said:
Yes, 100% agree. You'll want the inductor to be part of a high(ish) Q tank. This will help a lot with energy storage (efficiency) in the transmitter field as well as reduce EMI in the harmonics. The EMI from the drive circuit can be filtered and contained, but you don't want anything "extra" in the transmitter coil.Baluncore said:
It's not really your circuit question, and you probably don't have much control over the specs. But the coupling efficiency is a huge issue. Anything you can do to reduce the distance or increase the size of the coils will pay off. Can you add a magnetic core to better direct the flux? I'm thinking something like a pot core with a really, really, big gap.Sterer said:
For efficient inductive coupling, the wireless charging transmitter coil needs to be bigger than the range between the transmitter and the receiver coil. You say the range will be greater than 1 cm, but I do not see the diameter of the receiver coil.Sterer said:
I have aimed at sending a signal at around 275kHz. It has an LC tank where C is 54nF.berkeman said:
The number of turns will determine the transformer voltage ratio. The transmitter and receiver coils should be of similar dimensions. For an air-core inductor, the area will also change the voltage ratio.Sterer said:
I have three of the same coil, i am planning on stacking the two for the receiving side. It will be less efficient and more bulky, but i think its better than manually making a coil as that could turn out more inefficient.Baluncore said:
This is the best advice.DaveE said: