I had an error in all previous calculations, i don't know why did i divide by 10. The power calculated in the iteration is in mW, the time is in ns, so when multiplying that by time and dividing it by 1000000 (million) instead of billion, we get energy in uJ, and when multiplying that by frequency in Hz, we get power in uW. This is how it must be and this is how it now is. So all results were evidently 10 times less, but this didn't change the output/input power ratio. The Python calculation script with that error corrected, is now the following.

import sys

#Time scale in us
XU = 2.0
#Voltage scale in V
YU = 0.225
#Resistors 2 and 3 resistances in ohms
R2 = 987.2
R3 = 99.91
#Voltage of the power supply in V
V = 12.0
#Collector voltage in V
VC = 0.3
#Frequency in kHz
F = 110.001

#Oscilloscope scale is 100 squares, square is 10 units
#In calculations XU and YU are in ns and mV for unit
#Voltages are in mV and frequency is in Hz
V *= 1000
VC *= 1000
F *= 1000
ilist = []
olist = []
e = 0.0
n = 0

def gety(list, n, t):
    lx = float(list[n][2] - list[n][0])
    ly = float(list[n][3] - list[n][1])
    dx = float(t - list[n][0])
    dy = ly * abs(dx / lx)
    return (list[n][1] + dy) * YU

def output(s, e):
    #Energy in uJ
    e *= XU / 1000000
    print("%s power was %.3f uW" % (s, F * e))

for s in sys.stdin:
    l = []
    i1 = 1
    if (s[0] != "I" and s[0] != "O"): continue
    for j in range(4):
        i0 = i1 + 1
        i1 = s.find(" ", i0)
        if (i1 == -1): i1 = len(s)
        l.append(int(s[i0:i1]))
    if (s[0] == "I"): ilist.append(l)
    if (s[0] == "O"): olist.append(l)

for t in range(ilist[-1][2]):
    while (t >= ilist[n][2]): n += 1;
    vlr = gety(ilist, n, t)
    pr = vlr * vlr / R3 / 1000
    ilr = (V - VC - vlr) / R2
    plr = ilr * vlr / 1000
    pl = plr - pr
    e += pl
output("Input", e)

n = 0
e = 0.0
for t in range(olist[-1][2]):
    while (t >= olist[n][2]): n += 1;
    vlr = gety(olist, n, t)
    pr = vlr * vlr / R3 / 1000
    e += pr
output("Output", e)

The output of that script with the data from the previous coaxial coil test graph, was the following.

Input power was 173.492 uW
Output power was 79.925 uW

I also got the sample data of the same test directly from the oscilloscope screen image, as i described above. This sample data was the following.

https://paste.ubuntu.com/p/4nQzyT47W5

The following is the Python script that calculates directly from sample data, this can also be used to calculate from the sample data taken from the oscilloscope, provided that the numbers precede by the letter I or O. In that sample data one oscilloscope scale was 50 (which was also 50 pixels on screen), the numbers are relative to the x axis, and the output part is inverted.

import sys

#Time scale in us
XU = 2.0
#Voltage scale in V
YU = 0.225
#Resistors 2 and 3 resistances in ohms
R2 = 987.2
R3 = 99.91
#Voltage of the power supply in V
V = 12.0
#Collector voltage in V
VC = 0.3
#Frequency in kHz
F = 110.001

#The length of a scale is 50 pixels
#In calculations XU and YU are in ns and mV for unit
#Voltages are in mV and frequency is in Hz
V *= 1000
VC *= 1000
F *= 1000
XU *= 1000 / 50
YU *= 1000 / 50
ilist = []
olist = []
e = 0.0

for s in sys.stdin:
    if (s[0] == "I"): ilist.append(int(s[2 : len(s)]))
    if (s[0] == "O"): olist.append(int(s[2 : len(s)]))

for t in range(len(ilist)):
    vlr = ilist[t] * YU
    pr = vlr * vlr / R3 / 1000
    ilr = (V - VC - vlr) / R2
    plr = ilr * vlr / 1000
    pl = plr - pr
    e += pl
e *= XU / 1000000
print("Input power was %.3f uW" % (F * e))

e = 0.0
for t in range(len(olist)):
    vlr = olist[t] * YU
    pr = vlr * vlr / R3 / 1000
    e += pr
e *= XU / 1000000
print("Output power was %.3f uW" % (F * e))

The output of that Python script with the sample data above, was the following.

Input power was 209.347 uW
Output power was 111.954 uW

As one can see, this differs a lot from the results above. This shows the great imprecision of the method of extracting values directly from the oscilloscope's screen image.

If this Python script was still too complicated, split it into two as the following, calculating the input and output parts separately. Then you don't need letters before the numbers, just a list of the input part, and the list of the output part. List like this:

7
8
4

import sys

#Time scale in us
XU = 2.0
#Voltage scale in V
YU = 0.225
#Resistors 2 and 3 resistances in ohms
R2 = 987.2
R3 = 99.91
#Voltage of the power supply in V
V = 12.0
#Collector voltage in V
VC = 0.3
#Frequency in kHz
F = 110.001

#The length of a scale is 50 pixels
#In calculations XU and YU are in ns and mV for unit
#Voltages are in mV and frequency is in Hz
V *= 1000
VC *= 1000
F *= 1000
XU *= 1000 / 50
YU *= 1000 / 50
ilist = []
e = 0.0

for s in sys.stdin: ilist.append(int(s))

for t in range(len(ilist)):
    vlr = ilist[t] * YU
    pr = vlr * vlr / R3 / 1000
    ilr = (V - VC - vlr) / R2
    plr = ilr * vlr / 1000
    pl = plr - pr
    e += pl
e *= XU / 1000000
print("Input power was %.3f uW" % (F * e))

import sys

#Time scale in us
XU = 2.0
#Voltage scale in V
YU = 0.225
#Resistors 2 and 3 resistances in ohms
R2 = 987.2
R3 = 99.91
#Voltage of the power supply in V
V = 12.0
#Collector voltage in V
VC = 0.3
#Frequency in kHz
F = 110.001

#The length of a scale is 50 pixels
#In calculations XU and YU are in ns and mV for unit
#Voltages are in mV and frequency is in Hz
V *= 1000
VC *= 1000
F *= 1000
XU *= 1000 / 50
YU *= 1000 / 50
olist = []
e = 0.0

for s in sys.stdin: olist.append(int(s))

for t in range(len(olist)):
    vlr = olist[t] * YU
    pr = vlr * vlr / R3 / 1000
    e += pr
e *= XU / 1000000
print("Output power was %.3f uW" % (F * e))

Also if you have no other graphing tools, then with the following you can plot a graph from a list, a point graph or bar graph. Just paste the list to the column A there, and select A as the Y source. Then you can separate the input and output parts.

https://plot.ly/create/#/

Can this be self looped and drive a joule thief to indicate it's running?

Hi everybody

The current experimentations here and there with bifilar coils made me want to try it too (thanks guys). If anomalous results are observed, which remains to be confirmed, then we may have a theory behind it with nuclear or electronic spin resonance phenomena.  The NMR/ESR is the guiding idea that gets me started in experimentation, that's why I plan to change from conventional wire coils to flat copper tape coils. In spite it will not be an exact duplication of ayeaye's experiment, I would like to inform you that I started today. I failed to do a correct winding, the copper tape is very thin and easily damaged. I just ordered some of these tapes again, like these:
https://www.amazon.com/Double-Sided-Conductive-Shielding-Soldering-Electrical/dp/B074QN1YNH/ref=sr_1_25

These tapes should provide a higher capacity between turns. Although they are very thin, the effective conduction depth due to the skin effect is only 65µm at 1 MHz and 20 µm at 10MHz so we will not lose much of the effective resistance at RF frequencies.
I think another interest is the much better regularity of the electric field between the two tapes because of a constant spacing between conductors unlike wires of round cross-section.
The thinness of the tapes is also an advantage if nuclear or electronic spin resonance phenomena are involved, as there will be less dispersion of the magnetic field in the conductor bulk.
I hope for my first results next week (positive or negative!).

Spreadsheet of the input part.

Someone said maybe increasing R3 may help the coil to get more energy or such, i don't know...

I don't know what ones programming skills are, but below we will do it like spreadsheet, but not with spreadsheet.

In linux, there is a command line program named bc, that stands for "basic calculator". It is an arbitrary precision calculator, that is, it can calculate with any number of digits after the decimal point. By default, that is with the option -l, the number of digits is 20. It looks somehow primitive, and one may not believe it, but it is actually used by scientists. There is also the following windows version of it.

http://gnuwin32.sourceforge.net/packages/bc.htm

In windows, one should use the cmd console. For that in the windows explorer, go to the directory where your files are, then click on the address bar above, write cmd and press enter, this opens the cmd console in that directory. Selecting things and pressing enter will copy, and right click will paste there.

It's a problem to get a list into bc, as it wants things in the form like a[7] = 17. So, write the following script, that is create a text file and paste the following into it, like with wordpad. Maybe one can use bc also with some ide, but i don't know which one some may like. Then save it as tobc.bc.

for (j = 0; j < 134; j++) {
    n = read()
    print "a[", j, "] = ", n, "\n"
}

Provided that we have list of samples in the file list.txt, and that we also know that there are 134 numbers (again, we do calculations like in a spreadsheet), and that we have bc installed, write the following on the cmd console (followed by enter).

bc -l tobc.bc < list.txt > tobc.txt

Then write the following script that does the calculations, as you see, it calculates almost exactly like a spreadsheet. Save it as output.bc.

for (j = 0; j < 134; j++) b[j] = a[j]*0.225*1000/50
for (j = 0; j < 134; j++) c[j] = b[j]*b[j]/99.91/1000
for (j = 0; j < 134; j++) e[0] = e[0]+c[j]
e[1] = e[0]*2.0*1000/50/1000000
e[2] = 110.001*1000*e[1]
print e[0], "\n"
print e[1], "\n"
print e[2], "\n\n"
for (j = 0; j < 134; j++) print a[j], "  ", b[j], "  ", c[j], "\n"

Then write the following on the cmd console.

bc -l tobc.txt < output.bc > output.txt

The output is a bit primitive, as the bc language doesn't have the fancy things to write numbers in some format. The following is the output.txt, calculated for the output part like in the previous posts here.

https://paste.ubuntu.com/p/w5gBZkqtfV

The beginning of that output was the following.

25.44393954559103192763
.00101775758182364127
111.95435175818236334127

As one can see, the python result was correct, and the spreadsheet result was less precise, as i also said earlier. In fact, the python results were satisfactory. One can use spreadsheet as well when the precision is satisfactory, just that it has been proven now that spreadsheets are less precise than python, and in the scientific calculations they may in some cases not be satisfactory.

PS  For plotting, as one has to find that magic number, in whatever way, where the input part ends and the output part starts, to then split it into two lists of samples, for separate calculations of the input and output part. Other than doing it online as i did show earlier, one can also install the mingw version of gnuplot for windows, from the following link. There is a huge number of various plotting programs, whatever one may prefer, but gnuplot is the most classic.

https://sourceforge.net/projects/gnuplot/files/gnuplot/5.2.4

When there is a list of samples in the file list.txt, then run gnuplot and write  plot "list.txt"  followed by enter, this will plot a graph of it, it's that simple.

P = V * I

This is true about any part of the circuit, no matter whether there are coils or no coils. Both for input and output power of that part of the circuit.

Now it may look like that when the coil conducts more, the power consumption of the whole circuit may look like more. Thus one may think measuring power consumed by the coil by a dissipation of power on a resistor in series with the coil, etc. This does not measure the power consumed by the coil, not what concerns physics, different from some possible practical purposes.

Why? First because this is how power is always calculated, this is how electric power is measured, any electric power. Because this is the physical meaning of electric power. Several people who here doubted whether it's right, are therefore wrong. They are wrong in terms of physics, and this is what matters for research. They may not be wrong what concerns how some or other circuits are made, but this is not about the physical meaning of power.

So ok for their particular circuits, when more current goes through the coil, then their circuit consumes more power. But this is only about how their circuit is made, it is about their circuit and not about the physical power consumed by the coil. How so? Simply speaking, the current that goes through the coil, is not consumed by the coil. Like when the coil had zero resistance at any moment of time, the power it did consume at that moment of time were zero, zero and not a bit more.

Because again simply speaking, the current that goes through the coil, is most often dissipated in some resistance, and released as heat energy, thus none of that energy is lost in the coil. How the current that goes through the coil is used in the circuit, that depends on the circuit, whether it is used somehow or just wasted.

But the purpose of this experiment is to measure the input and output power of the coil, coil and coil only. Not the power in some circuit that can be incredible wasteful as well. But how much power some wasteful circuit consumes, gives no useful information about the coil, none and nothing, thus it makes no sense to measure, and has no physical significance whatsoever about the coil.

Please understand these things first. It makes no sense to do the experiments to measure the power efficiency of a coil otherwise.

You may want to review this PDF slideshow on power measurement, from IEEE:

http://ieee.rackoneup.net/rrvs/13/fundpwrmea.pdf

First of all, in this experiment the power is not measured by whole waveforms, so what at all has this to do with it?

This pdf is correct, but it is about something different. If you disagree and if you insist, i will talk directly to the IEEE guy, i'm sure he does understand.

But no matter what, you may notice even from that document, power is always  P = V * I  at any moment of time, this doesn't depend on what circuit or a part of it is measured.

If we measure power at an instant of time, then it's always  P = V * I . If we don't measure power at an instant of time, but measure like the amplitude of a sine waveform instead, then we have to multiply it by cosine fii, this comes from the waveform, not that power is any different. And even when we measure sine waves, no matter what circuit or part of circuit we measure, no matter whether it contains coils or contains no coils, power is always measured the same way, not different when it is a coil or the circuit or a part of it contains a coil, as you insist. It is never said in that pdf either, that a different equipment or a method of measurement should be used depending whether the part of circuit measured contains a coil or not, read by yourself if you don't believe.

P = I * V is not correct if the current and voltage are not in phase.

Holy cow, phase is only important when we measure a whole waveform, like when we measure the amplitude of a sine waveform and know that the form of the signal is a sine. But in this experiment we measure the power at every *instant* of time, during the waveform, not the whole waveform. And at every instant of time  P = V * I , at every instant of time, this is not true like for the amplitude of the sine waveform during all its duration.

I know what you are talking about, i learned about it in the university too. But i didn't learn electrical engineering, i learned automatic control, and there is often important to do measurement at every instant of time, and do calculations from it. Say you regulate a dc power voltage, it can change in every kind of ways, sure you cannot measure power by sine or any waveform, this and everything else has to be calculated at every instant of time.

To show you how the power calculated from the amplitudes of sine, comes from the instantaneous power, let's calculate a power of a sine signal, with amplitude 1, so that the voltage and power are shifted by 20 degrees. Using the following bc script, you can try it. We calculate it during a time 10000 seconds, and then calculate the same for the sine waveforms by  Vrms * Irms * cos(20 degrees) . The amplitude of both our voltage and current is 1, but by that equation both have to be root mean square, so it becomes  rms * rms * cos(20 degrees) .

pi = 3.1415926535897932384626433
pi2 = 2 * pi
#20 degrees to radians
r20 = 20 / 360 * pi2
rms = 1 / sqrt(2)
energy = 0
for (t = 0; t < 10000; t += 0.1) {
    current = s(t)
    voltage = s(t + r20)
    power = voltage * current
    energy += power
}
energy *= 0.1
power = energy / 10000
print "Power when measured at instants of time: ", power, "\n"
power = rms * rms * c(r20)
print "Power when calculated from the amplitudes of sine: ", power, "\n"
quit

The output of that script is the following. As you see, the results are quite close, and the results become closer the longer is the time over which we calculate it.

Power when measured at instants of time: .46983333927318779782
Power when calculated from the amplitudes of sine: .4698463103929541\
9202

So as you see, the instantaneous  P = V * I  is correct, different from that said.