Kapanadze Cousin - DALLY FREE ENERGY

[verpies]

Why a scope with carefully isolated probe cannot sense well enough, I cannot understand the need ?

An average scope can sense the voltage well enough. 
An average 2-ch scope can sense the voltage and current well enough with an aid of a CSR.

...BUT an average scope often cannot:
a) sample 2 channels simultaneously, in order to create two time-synchronized signals.
b) multiply these 2 signals (instantaneous voltage and current), with decent precision.
c) average results of these multiplications (products).

There are scopes that can simultaneously sample, multiply and then average 2 channels at the full bandwidth of their ADCs, but such scopes are rare and very expensive.  Also digital scopes run into additional problems of multiplying two quantization errors of their ADCs, especially when improperly amplified input signals, toggle only few LSBs of their ADCs.  For an example of this ugliness see here and here.

@Itsu
Please show ARMCORTEX that video illustrating what happens when you try to multiply 2 channels with your scope.

Where is useability and worth of such circuit Verpies ?

It allows a poor man to measure DC - HF AC power quickly and accurately and display it on a voltmeter.
Two such circuits and two voltmeters will allow an experimenter to spot any O/I power anomaly^* immediately and accurately (even if it lasted only 1s), without the need to own two very expensive scopes and dealing with the the nasty issue of multiplying two quantization errors when using DSOs.

Adding a microcontroller to such circuit would allow for long-term integration of power, which would yield an energy measurement.
As a small additional bonus, a microcontroller and an LCD would eliminate the need to tie up a voltmeter to act as a display.

Because the 2GHz multiplier costs less than $10 and the rest of the components are not much more expensive, and two such circuits allow for the real time monitoring of an O/I power ratio - I deem the worth and usability of this circuit as great for the members of this forum and for anyone else who wants to display DC - HF AC power delivery (or consumption) on his voltmeter.

* Or lack thereof

[d3x0r]

An average scope can sense the voltage well enough. 
An average 2-ch scope can sense the voltage and current well enough with and aid of a CSR.
...BUT an average scope cannot multiply 2-ch (instantaneous voltage and current), well enough and often it cannot average results of these multiplications at all.

There are scopes that can multiply and average 2 channels at the full bandwidth of their ADCs but they are very expensive and digital scopes run into additional problems of multiplying two quantization errors of their ADCs, especially when improperly amplified input signals toggle only few LSBs of their ADCs.  For an example of this ugliness see here.
@Itsu
Please show him your video, when you try to do that with your scope.
It allows a poor man to measure DC- HF AC power quickly and accurately and display it on a voltmeter.
Two such circuits and two voltmeters will allow an experimenter to spot any O/I power anomaly^* immediately and accurately (even if it lasted only 1s), without the need to own two very expensive scopes and dealing with the the nasty issue of multiplying two quantization errors, in case of using DSOs.

Adding a microcontroller to such circuit would allow for long-term integration of power, which would yield an energy measurement.
As a small additional bonus, a microcontroller and an LCD would eliminate the need for a voltmeter acting as a display.

Because the 2GHz multiplier costs less than $10 and the rest of the components are not much more expensive, and two such circuits allow for the real time monitoring of an O/I power ratio - I deem the usability of this circuit as great for the members of this forum and for anyone lese who wants to display AC power delivery or consumption on his voltmeter.

* Or lack thereof

how much power will it withstand?

[verpies]

how much power will it withstand?

Did you look at the schematic at all?!

All of the power flows through R1, only a small proportion of it is diverted for measurement.
So, the direct answer to your question is: "As much power as R1 can handle".

[d3x0r]

Did you look at the schematic at all?!

All of the power flows through R1, only a small proportion of it is diverted for measurement.
So, the direct answer to your question is: "As much power as R1 can handle".

that was my suspicion, thanx for confirming it.  Making sure I wasn't missing some other point.

[verpies]

The additional suggestions would also be of great help and maybe a current limit control to stop the FET's from smoking too !! 

Of course limiting the drain current on a pulse-by-pulse basis can be added to the Mazzilli circuit, which together with gate overvoltage protection would prevent the FET "smoking".  Pulse-by-pulse current limiting is already implemented in many PWM Controller ICs.  I anticipate a little problem with the Miller effect when high dv/dt is involved but it can be remedied with various techniques.

In any case, a robust resonant LC driver can be made, but it will not be as simple as the 2 transistor Mazzilli circuit anymore.
Will minimalistic experimenters, like NickZ, want to sacrifice simplicity for robustness ? - that is the question !

As the complexity of the circuit increases, another question comes up. 
Namely, what are the desired frequency determining components ?

Do you want the internal LC components and a load to determine the oscillation frequency,
or do you want this frequency to be determined by external and load-independent components ?
e.g. the venerable 555 timer or your precise benchtop signal generator.  If it's the latter then see this.

The Mazzilli circuit essentially is a digital Load Controlled Oscillator   In some applications this behavior might be desirable and in some - it might not.

P.S.
A digitally switching LC driver does not generate a pure sine wave in its LC circuit, but it is more powerful than an analog driver costing the same. 
FYI: The original Mazzilli circuit switches more digitally than a true analog LC driver. 
Nonetheless a conscious decision should be made from the start, whether a digital or analog driver is needed (e.g. because of spectral purity issues...)