Given this thread has resurfaced I'll try to explain what the guys of the 'QEG generator' think they are up to, and why they epically fail.
The big problem is that human intuition is badly equipped to cope with alternative currents (on a side note, another topic where human intuition fails badly is with probabilities). Indeed, the topic of power with alternate currents is apparently simple, but actually difficult even for engineers (I'm an electronic engineer, by the way, and I had my share of problems when I approached it).
Many people know the formula power = voltage * current, but what most people miss is that the formula is only valid for continuous currents. With alternative currents the formula changes: power = voltage * current * cos(phi), where cos(phi) is the cosine of the phase angle between the (sine wave, for simplicity) voltage and the (sine-wave) current.
[Another side note: in the continuos case, voltage and current are easily defined concepts. In the alternative case things are not so easy and there are at least three different meanings: 'average' voltage/current, 'peak' voltage/current and 'root-mean-square' (RMS) voltage/current. The power formula above must use the RMS values. A normal tester measures RMS values]
Let's go back to W = V*I*cos(phi). Forgetting the cos(phi) is terrible: you can have a circuit with 100KV and 100KA (as measured by a tester) flowing through it and.. exactly zero power. You just need to have the voltage 90 degrees out-of-phase with the current and cos(phi) becomes zero.
In the electrical jargon the quantity V*I is called 'apparent power', the quantity V*I*cos(phi) is called 'active power' and what is left (V*I - V*I*cos(phi)) is called 'reactive power'. 'Reactive power' is a bad misnomer, because it's not a power at all: it can do no work and it would be better named 'fake power'. It's just electrons shuttling back and forth, adding to the (inevitable) resistive losses in the circuit but giving out nothing useful. By the way, this is the reason why electrical utilities require the cos(phi) to be above a certain value: because they don't want to have losses which are not paid for (the electrical power meters used for billing only measure active power).
And now let's see the 'QEG generator'.
Question: is it possible to build a circuit which takes as input an alternative current of say 100V 1A and gives out say 100kV and 100kA? Oh yes it's quite possibile (at least in theory, you'll require some pretty high-end technology to reach 100kV and 100kA, but that's not the matter). One uses a resonant (inductor-capacitor, or 'LC') circuit which is suitably 'pumped' from the input. It's just the same as a see-saw: with small pushes you can make a see-saw swing a lot.
Question: having transformed 100V 1A (let's say 100W, cos(phi) = 1) into 100kV and 100kA, can we now use those 100kV 100kA to deliver a stunning ten thousand megawatts (100K * 100K)? Not at all! The cos(phi) in a perfect LC circuit is.. guess what.. zero, and your net power output will be zero too. And where is the 100W input power going, then? It's used to overcome the (inevitable) circuital losses, of course, just as, in the see-saw example, part of the energy of the pushes is used to overcome pivot and air friction.
What if we add a resistor in the circuit, so it will heat and we can extract energy from? This will make the LC resonant circuit less ideal and the cos(phi) will increase, so we are now actually extracting power from the circuit. But where does this power come from? Well.. from our input.. now we have bigger losses to overcome and our input will draw more power to compensate, just as adding a brake on the pivot of the see-saw will require us to push harder.
From an engineer point of view the 'QEG generator' is a pumped resonant circuit. Judging from the diagrams and the pictures.. it's extremely inefficient, both for the construction of the LC circuit itself and for the naif way used to pump it. By 'inefficient' I mean that only a minimal fraction of the input power (my bet: less than 5% and possibly much less) is being stored in the resonant circuit (up to a point), all the rest are losses dissipated to heat. It also looks very dangerous.. I'd steer at least 20 meters clear from any appliance with working voltages in the thousands and the appearance of junkyard trash as shown in the pictures. Electrocution, fire and even explosion hazards look substantial.
Last remark: they also make a mess with the meaurements. As someone already said they feed a peak-to-peak measure (taken with an oscilloscope) into a formula where an RMS measure must be used (and the difference is quite significant). I'd immediately fire any engineer who makes an error like this, but this is not the root cause of the failure of the 'QEG generator'. Ah, and what 'quanta' have to do with all this, is anybody's guess.