Wireless Power transmission. Is it possible to achieve 100-200w (load
side) power using off-the shelf components?
Here's an answer that takes you some distance in understanding some of the physics and the problems. What you want sounds a lot like how a transformer works but without the magnetic core shared by the primary and secondary windings. For a regular transformer you can easily get power transfer efficiencies greater than 95% but without a core you start to hit problems.
The main problem is that the alternating magnetic flux produced by the driven coil only loosely couples with the receiving coil. At best, given the scenario described, this might be 30%. With only 30% of the flux received, the receiver has to have 3 times as many turns to get a 1:1 voltage transfer. That's a voltage transfer and not a power transfer.
When you have a load connected that takes significant current, that field starts to reduce - the current in the receive coil produces a counteracting field. This then reduces the net field and voltage drops. More load current means more voltage reductions.
To counter this, designers use resonance. Resonating the primary transmit winding with a capacitor means that the nature of the current flowing in the primary is massively increased (maybe ten or twenty fold). Because there is more current there is more magnetic flux proportionately.
To get 100 watts across a gap probably means a current in the realm of several tens of amps RMS and this starts to mean you need to use Litz wire (easy to use and very neat but rather expensive).
It also means you need a fairly stable high frequency power oscillator that is capable of shoving out 100 watts plus all the power lost. This might be in the realm of 200 to 300 watts. Without a load on the receiver that's not a big deal but you have to be wary of producing emissions that can effect other local pieces of electronics.
All in all it's a problematic job for someone experienced in electronics and magnetics. I could probably crack it after a few months (and I have patents in this sort of AC magnetic technology) but it might take you a year or so or even longer.
Here's a simulation: -
I've chosen a primary inductance of 1uH, a coupling of 50% and a secondary inductance of 4 uH (twice the number of turns as primary). There is a 1 ohm load connected to the secondary and that load is series tuned with 8.4 nF to maximize power transfer. The oscillator runs at 1 MHz
When you do the math to calculate the capacitor value, you find that the effective inductance after all the partial coupling is 3 uH.
With Rval (primary coil resistance) of 0.1 ohms, it will be twice this in the secondary because there are twice the number of turns. To get 100 watts out requires 28.8 Vp-p across the 1 ohm load.
This requires 37.2 Vp-p at the input and an input current of about 10 amps RMS. My simulator tells me that the input power is about 131 watts so that's not a bad inefficiency but remember that coil coupling is 50% and overly optimistic.
So now, the coupling has dropped to 30% and immediately we have lost resonant tuning because the effective value of the inductance has changed. With the previous input being the same, the output power is a measly 5.4 Vp-p. To restore the output to being 100 watts now needs an input power of about 150 watts but now the input voltage has to be massively increased from 37.2 Vp-p to 194 Vp-p to overcome all the de-tuning effect of moving the coupling (this is the same as moving the two coils apart a little bit).
Do you see the logistics of the problem? You can optimize the tuning to give a decent power efficiency at a certain distance then, when you move the coils apart you get massive de-tuning and you then have to force a lot of reactive power into the primary winding to get what you want in terms of output power.
The simulation is a very simple circuit - there is no power oscillator and there is no receiver rectification and regulation - it's just a sine wave in and sine wave out across a 1 ohm load.