Sounds like an interesting setup however I can see what MH is saying.   Right now it seems you are measuring the power dissipation across a resistor on each side of the transformer but that does not mean it's the total power used in the circuit.  I believe you would see more power measuring across the input from the sig gen as well as more power at the output across the transformer.   If that output is still higher than input I believe you have something to get excited about.      But that's just my thoughts -- I think both you and MH are way ahead of me on theory.   

Tinman:

From the voltage across a resistance you can measure the power dissipated in the resistance itself.  I can't be sure, but you seem to be indicating that you can measure the input power to the circuit like that, but you can't.

Note by measuring the voltage across a resistance you are also measuring the current through the resistance.  In looking at your first circuit you have two current loops and a bunch of known resistances in those loops.  So you have enough information available to measure the power dissipated in each individual resistive component in the circuit, and thus the total resistive power dissipation of the entire circuit.

The next step would be to measure the input power to the circuit as supplied by the signal generator and compare that to the total power dissipated by the circuit.

MileHigh

P.S.:  This is a bit more advanced:  Let's say you now know how much power is being dissipated by the entire circuit.  You also know how much current is flowing in the signal generator current loop.  Then you measure the signal generator voltage.

In theory that gives you enough information to determine the phase angle between the voltage and the current supplied by the signal generator, no scope required.  So if you make good measurements, you should be able to hook up your scope with an expectation of what the phase angle should be before you even see it.

Just for fun let's try going one step further.

I am going to make some assumptions about two other parameters.  I don't know the wire lengths in the setup.  All that I know is the signal generator frequency is close to one megahertz so I will take a guess that there are some observable inductance effects.  That would show up as a lagging phase angle in the signal generator current compared to the signal generator voltage.  I am also going to assume that there are measurable losses in the core.  Of course there are losses in the core, but are they measurable or not is the question.

So the goal is to measure the losses in the core by inference from other measurements.  You have an accurate measurement of the total resistive power dissipation of the circuit.  You try to measure the signal generator voltage and current and phase angle as accurately and carefully as possible.  Here is where you play with the horizontal and vertical stretching on your scope display to line up a single or half sine wave with the display graticule.  That makes it easy to make precise phase angle measurements.

Your real power measurement with your scope is Vrms x Irms x Cos theta.   Compare that with the total resistive power dissipation of the circuit.  If the scope-measured real power is greater than the total resistive power then you can infer that the "extra" power measured by the scope method is due to the heat power losses in the core.

Then if you run your circuit for say ten minutes and compare the before/after core temperatures you might get basic empirical confirmation that the core is indeed heating up.  If you just can't stop taking extra steps, you could measure the rate of core temperature increase and knowing the mass and specific heat of the core material, measure the heat power in the core that way.  Not too difficult to make slight physical changes to the bifilar coil so the core itself is in a thermal tomb.  The coil would not be affected electrically.  Then compare that core heat measurement with your derived core heat power from the scope measurement.  That might be something that's right up TK's alley.  It's like a vivisection!!!

See now that would be fun for me!  The funny thing is I never made these kinds of measurements in real life.  lol

MileHigh

Last gasp....

How can you get the most accurate total thermal power dissipation measurement?

You know that you have quite accurate current measurements for each loop, which is a good start.  So with your multimeter, you measure the electrical resistance for the entire loop, i.e.; all the components in series.  So you make only one resistance measurement with your multimeter for each loop, which is much more accurate than adding individual measured resistances together.  Note this also factors in the total wire resistance in each loop.

This helps in the quest to measure the power dissipation in the core by inference.

Quote from: MileHigh on August 31, 2013, 06:27:46 PM
Last gasp....

How can you get the most accurate total thermal power dissipation measurement?

You know that you have quite accurate current measurements for each loop, which is a good start.  So with your multimeter, you measure the electrical resistance for the entire loop, i.e.; all the components in series.  So you make only one resistance measurement with your multimeter for each loop, which is much more accurate than adding individual measured resistances together.  Note this also factors in the total wire resistance in each loop.

This helps in the quest to measure the power dissipation in the core by inference.

Mmm-a couple of problems here.
The resistance of L1 will go up with frequency,along with L2,due to the skin effect.
Second problem-my DMM's are usless at these frequencies,as they read 0 on amp's ,and voltage is all over the place-both AC and DC setting.
So i will have to use the scope to measure across the resistor,then also across the coil for voltage.

As the power level's are very low here,the core dose not change in temp that i can detect with my IR temp gun,not even .2*f-after 30 minutes running.