1

I am designing a wireless power transfer system based on the Royer oscillator ZVS principle. I based my circuit on the following image, this can be found in the application note [ANP032] page 21. enter image description here

The circuit I have designed is given below. enter image description here

The difference in this design is the following components: D701, D703 and D702 are 1N4148WS diodes with a If of 150 mA; Diodes D705, D704 and D700 are TVS bidirectional diodes 824521601.

The diodes given in the Example are: D701-> 1N5819HW: Io=1 A; D702 & D703-> Rb168M150: Io=1 A; D704 & D705-> Unidirectional TVS diodes 824550781; D700-> Bidirectional TVS diodes 824541151

I noticed that with whatever combination of coils, or capacitors I use for the Resonant frequency, I am only able to read about 0.1 A at the receiver end, with varying voltage levels (depending on the input, coil used for the receiver coil). I have a 200 Ω load on the receiver end. Also with increased current i.e., 2 A on the Tx side, the diode D705 always burns out. My observation was that I may need to change the diodes mainly D701, D705 and D706, but I am just shooting in the dark. Would this be the right assumption or is their something else wrong with the circuit?

The receiver circuit. J600 and J601 are both two pin connectors. enter image description here

Edit 1: enter image description here

EmeraldMonk
  • 175
  • 8
  • Why did you use MOSFETs instead of diodes in your receiver? Did you tune the receiver to peak performance? – Andy aka Jan 20 '23 at 14:55
  • I made two receiver circuits, one with diodes for the full bridge rectifier, and the one with the MOSFETs was the second one i found, thought to make it for testing purposes. The circuit with the diodes burnt through, and this was the one that lasted with varying voltages. – EmeraldMonk Jan 20 '23 at 15:04
  • Did you monitor the output DC voltage when the diodes burned? This usually happens because you haven't implemented a voltage regulator for your load. What TVS diode did you use? Did you notice that your LED is the wrong way round? – Andy aka Jan 20 '23 at 15:17
  • I do not have a voltage regulator on the receiver side, this is because I am not aware of the voltage/current to expect. I gave an input of 24V,(Not sure of the exact current input <1.2A) ; measured a voltage somewhere close to 70V. – EmeraldMonk Jan 20 '23 at 15:25
  • I have slightly detuned the receiver circuit resonance frequency by 15-20% when compared to the Transmitter circuit. – EmeraldMonk Jan 20 '23 at 15:27
  • I used the TVS diode 824541151. Yes I realised that after a few tests, I ended up removing the LED. – EmeraldMonk Jan 20 '23 at 15:31
  • *measured a voltage somewhere close to 70V* <-- I think the TVS diode is rated for 60 volts so, it will heavily clamp at 70 volts and, that could be a very significant reason for diode failure. Your capacitors are also only rated at 50 volts so you have definitely exceeded the limits to which the design is capable of. – Andy aka Jan 20 '23 at 15:32
  • To maximize power transfer to a regulated output (say 5 volts) you should use a synchronous buck converter that is capable of dealing with an input DC voltage in excess of 60 to 100 volts. – Andy aka Jan 20 '23 at 15:34
  • I made the receiver circuit again with the diodes. Vin=15V,0.42A. Vout measured on the Rload on the Rx end-> 200ohm. 20V,0.1A. This has been a common theme with what I am able to achieve on the RX 0.1A with varying voltages depending on the L,C of the Rx and TX. anything more than that is at the cost of either too high of a voltage, or temperature. Would a synchronous buck converter help achieving a higher current output as well? or only regulate the voltage? – EmeraldMonk Jan 20 '23 at 16:04
  • Any buck converter will have a 90% power conversion efficiency so, if the input voltage is (say) 20 volts @ 100 mA the power is 2 watts. If the buck output voltage is 5 volts, you have 360 mA to play with, possibly a shade more with a good converter. If you choose a high voltage input buck converter you can gain more efficiency by more closely tuning the receiver to the transmitter. – Andy aka Jan 20 '23 at 16:08
  • I noticed that the reason for only getting 0.1A was because I used a 10 parallel resistors as a Load which can handle about 5W combined, which translates to 158mA for a 200ohm resistor load. When I connected the Rload to an [electronic load](https://www.gmc-instruments.de/produkte/spl-250-30) I was able to measure a much higher current. Some coils even reaching to 0.6A on the Rx end. – EmeraldMonk Jan 24 '23 at 08:30

1 Answers1

1

is there something else wrong with the circuit?

The receiver TVS diode is rated at 60 volts yet the electrolytic capacitors are only 50 volt type. The receive LED is incorrectly connected but this is trivial.

measured a voltage somewhere close to 70V

Then it's likely that this is the reason why your diodes burned; the TVS would heavily conduct and the diodes will "meet their maker" so to speak.

You should be benefiting from the fact that you have got such a decent received voltage and use a synchronous buck regulator to drop the voltage towards your target stabilized DC output (possibly 5 volts?). Use receiver tuning capacitors that are more than 100 volt rated and, pick a smoothing capacitor (after the bridge) that is also rated more than 100 volts.

I think your tuning capacitors are 250 volt rated so that's plenty good enough. However, do they have a decent dielectric like C0G. You can benefit from C0G capacitors a lot of the time.

I have slightly detuned the receiver circuit resonance frequency by 15-20% when compared to the Transmitter circuit.

That is a lot of detuning. With better tuning, you will benefit from a type of regulation (when the coils get close) because of the natural detuning effect as was shown in this answer. When significantly detuned, the effect isn't as good and, the distance-to-raw-output-voltage regulation can be horrible.

Andy aka
  • 434,556
  • 28
  • 351
  • 777
  • 1
    Yes, the tuning caps are C0G, whereas the smoothening caps are X7R. I will test out with a more closer resonance frequency of the Rx with the Tx. Hopefully I can get a better Pout/efficiency. – EmeraldMonk Jan 24 '23 at 08:25
  • I am aiming for 20V,0.2A at the Rx end. – EmeraldMonk Jan 24 '23 at 08:32