Quantum Energy Generator (QEG) Open Sourced (by HopeGirl)

[Farmhand]

Yeah it does change quickly with a frequency change of the primary because it's all tuned. I load the output and the primary max
voltage rises but the average voltage remains about the same, that's because the load changes the primary tune of course.
a load either increases primary power or decreases it depending on how I tune it and the load, I can tune the circuit to a load.
Or tune an inductive load to the circuit ect.

I can vary the input voltage, frequency, pulse width and vary the tank resonant frequency by adjusting the variable capacitor.
The tank and the load circuits are separate but connected circuits.

With this arrangement I can just change the pulse generator frequency and run the primary at a different frequency then tune
the tank to be resonant at or near the primary frequency. I can also tune the tank to double the primary frequency and still get
resonant rise.

..

[Farmhand]

Once you multiply  Vrms by Irms by the PF for your load, you get Watts. That is the real power P (AKA True Power),
and real power is expressed in Watts.
VA is used to represent the apparent power, signified as S. This would be your measured Vrms times your measured Irms for your load.
VAR is used to represent the reactive power, signified as Q
Apparent power is the vector sum of the reactive and real power.

That's exactly what I said. VA - VAR = Watts or Watts = VA - VAR or VAR + Watts = VA, same thing.
ie. 100 VA - 80 VAR leaves 20 Watts missing or dissipated. Fair enough ?

I do understand the concept. But I won't take offense. 
.

[Void]

Yeah it does change quickly with a frequency change of the primary because it's all tuned. I load the output and the primary max
voltage rises but the average voltage remains about the same, that's because the load changes the primary tune of course.
a load either increases primary power or decreases it depending on how I tune it and the load, I can tune the circuit to a load.
Or tune an inductive load to the circuit ect.

I can vary the input voltage, frequency, pulse width and vary the tank resonant frequency by adjusting the variable capacitor.
The tank and the load circuits are separate but connected circuits.

With this arrangement I can just change the pulse generator frequency and run the primary at a different frequency then tune
the tank to be resonant at or near the primary frequency. I can also tune the tank to double the primary frequency and still get
resonant rise.
..

Yes, OK. As I have mentioned, I have seen cases like this where I measure odd phase shifts like this, but I have
noticed that if I change the frequency a little bit so that the phase shift comes more into a normal range that there
doesn't appear to be any noticeable change in input and output power. If the tube gives off about the same
amount of light when the frequency is shifted a bit where the phase shift on the load current looks more reasonable,
then although I can't explain it, it would seem to be some just some weird phase effect which doesn't really accurately reflect the
actual load power.

[Void]

That's exactly what I said. VA - VAR = Watts or Watts = VA - VAR or VAR + Watts = VA, same thing.
ie. 100 VA - 80 VAR leaves 20 Watts missing or dissipated. Fair enough ?

I do understand the concept. But I won't take offense. 
.

Maybe it was just a typo then, but you wrote:
"83.5 VA x 0.987 = 78.4 VAR"
You also seem to be trying to subtract VAR's from VA's? You can't do that mate. 
They are vectors at different phase angles. This is why they all have different units.
You have to use special vector math to add and subtract vector values at different angles.
You can't use ordinary addition and subtraction.
Just Vrms x Irms x PF is all you need to calculate real power,
but of course these measurements have to be accurate.

Using your odd measured numbers, I get:
(A current lag of 200 degrees is the same as a current lead of 160 degrees, and both give the same power factor,
so the phase angle can be expressed as lagging 200 or leading 160 degrees. Either way, same result.)
84.6 Vrms x 0.98 Arms x Cos(160) =  -77.91 Watts.  Now that's impressive. 
(I am assuming you are using a 0.1 ohm CSR)

I can't explain it, but I will think about it. Maybe someone can see what is happening.
The difference in inductance introduced by a wire wound 0.1 ohm CSR and a non inductive CSR
should not make much difference here I wouldn't think, unless whatever you are using has
an unusually large amount of inductance.  Probably something else is going on here.

Edit:
Do you have your scope probes set to x10 and your scope channel settings configured for the same multiplication factor of x10 as well?
Just wondering if that might be a factor for the measured voltages? That should not affect phase angle measurement though.

P.S. For your 'betterflurosh.csv' file scope data, I calculated an average power of 2.597 Watts.
I assumed a CSR value of 0.1 ohms for the channel 2 current measurements.

[MarkE]

Most modestly priced scopes only offer a 20MHz filter.

For stable repetitive waveforms, a better way to remove noise artifacts on a digital scope is to apply averaging.  Averaging is usually found in the acquisition menu.  Averaging should be used with care.  Alwasy look at waveforms with averaging turned off before applying it.