Simple to build isolation transformer that consumes less power than it gives out

[Lynxsteam]

Can someone draw a diagram of how you think current flow is moving for one half of the sine wave?  I am unclear what the intended path is given all the various dots, wiring conventions, tags, etc....  I can see how the voltage could be amplified if the sine wave is in phase, but I don't see how amps wouldn't drop.  Also, in any cored transformer you lose some power due to eddy currents (heat).  And heat in the windings due to resistance.

[wattsup]

The watt meter most likely shows measurement error. In best case you could find 100W+ 1 Ohm resistor and attach to "cold" wire from mains in series then put oscilloscope on resistor and see RMS amps.
You are mixing reactive power coming back from transformers with conventional power so funny things happen.

@T-1000

Well, the problem is with this set-up that the output transformer E-F-G-H was starting to give some slight smoke so I stopped the video. I will start again tonight and make a clear connection diagram and do some tests at lower voltage off the Variac. But again, my diagram is good for my transformers and may not apply to others.

The info on my transformers was supplied to me by the manufacturer. I asked them how many wire turns per coil and winding direction per coils but they did not provide that information. Just the info sheet as shown.

The point here is by using the same @JN diagram, if you start switching the wires around and re-do your tests, you may get to the same situation as I am at now.

What is incredible is that I am getting more voltage output then the applied input voltage while still being at the 120 volts connection mode so there is something happening with this method. If this was making a simple short circuit, that would explain the input amps/voltage/watts, but 2000 watts seems so high for such a reading, and then try to explain the increased output voltage.

Something is funny and it is not Red Skelton (although he is very funny).

wattsup

[JouleSeeker]

@T-1000

Well, the problem is with this set-up that the output transformer E-F-G-H was starting to give some slight smoke so I stopped the video. I will start again tonight and make a clear connection diagram and do some tests at lower voltage off the Variac. ...
What is incredible is that I am getting more voltage output then the applied input voltage while still being at the 120 volts connection mode so there is something happening with this method. If this was making a simple short circuit, that would explain the input amps/voltage/watts, but 2000 watts seems so high for such a reading, and then try to explain the increased output voltage.

Something is funny and it is not Red Skelton (although he is very funny).

wattsup

Looking forward to seeing your connection diagram.  I'll try it.

Yes, the 2kW is indeed odd...  easiest thing to do would be to measure the input current using a clamp-on meter, as a check.  If things are starting to smoke, be quick 

Hey, I remember Red Skelton -- very funny guy.

[JouleSeeker]

  Having suggested a check using a clamp-on ammeter, I did it.. with interesting results.

See photo -- current shows as 0.12A, whereas the P3 kill-a-watt meter says about 0.2 A (and 19.5 W in). 
0.12A x 113.5V in = 13.6 W input.

The output P3 meter shows 64.5V and 15.8 W; but I'm beginning to suspect its accuracy here -- at least I need to check with different meters!  The clamp-on shows output current of 0.20 A (P3 meter shows 0.24A).
0.20A x 64.5V out = 12.9W output. 

At least we can compare with above, having used consistent tools:  12.9 out/13.6 in = 95%.
Which is quite good, since as Lynx noted, there are losses in the two trafos.
Indeed, using one of these trafos, I found its eff ~ 84%, for just one trafo.

BUT I'm still trying to convince myself of the measurements... how we measure Pin and Pout.

The main conclusion is -- I don't trust using just the P3 Kill-a-watt power meter (hereafter, P3 meter) alone... even though it tested out well using light bulbs of known Lumens, straight from the mains -- there everything agreed and gave me some confidence in these meters.

Also -- I have another clamp-on ammeter, checked that, and the two clamp-ons agree with each other very well.  (And disagree with the P3, especially for the input current measurement.)

[wattsup]

@JouleSeeker

Here is the diagram. Hope it is understandable.

I agree about the P3 as well as about mine. There is something not right with the reading. But what do you expect from a 20$ watt counting meter that is designed for 60Hz appliances that consume a steady and "non flybackable" (hichic) energy consumption. I think our regular measuring instruments have worked well till now. It is only when the amps are in the single decimal point range that we have to be careful and better to simply put a multimeter inline.

Anyways, the diagram shows exactly how I connected for the last video and I will do some more experimenting tonight. I will put two scope probes at various locations and see if there is any phase differences. I guess if the phasing is n ot the same, I should see two sine waves slightly offset.

Also, on the output, if this is AC output, then both output locations should show the same voltage level. Right or wrong? Well in my set-up, when you check the output on the scope on each of the load points, one is about many times higher then the other.  Don't know exactly how high yet. I will have to find some more scope probes because my last one just fizzed a few days ago. I'll have to dig around the office since i know I have two more somewhere. lol

Lastly, maybe @MH can see through this and come up with a half decent explanation but for now, I am at a loss to explain it.

wattsup