Sharing ideas on how to make a more efficent motor using Flyback (MODERATED)

[verpies]

is this 10% using just one coil, and could using ten coils you know what?

Average of several 10% is still 10%

[itsu]

Long      video here:  https://www.youtube.com/watch?v=dANzzD6X75E&feature=youtu.be

The stopping of the rotor while measuring the input is shown around 5:50 minutes.


Itsu

[verpies]

Hmmm,   I checked again the coil, and it measures 37.8mH @ 100Hz,

That's good to know but the coil's inductance is not necessary to make the input energy calculation when using the fool-proof, albeit difficult method of integrating V_INST * i_INST for the length of the ON-period.

the capacitor C2 measures 10.03uF @ 100Hz

So if the peak voltage across C2 is trustworthy then the energy recovered in C2 is trustworthy, too. All according to E₂=½C₂V²

and the CSR measures the mentioned 0.1 Ohm 1%, although that is hard to confirm as my DMM is fluctuating in that low region, but after subtracting the leads resistance, it seems to be the correct 0.1 Ohm.

Yes, it is difficult to measure low resistances.  The best way is to put exactly 2.000V across it and verify the the current flowing through it is exactly 20A since most current probes and multimeters can measure up to 20A.

The battery voltage was/is 12.4V as can be seen on the screenshot below where i took an input measurement of the battery voltage (12.4V yellow trace)

OK, So my previous calculation of the input energy per 1 pulse would be now: ½ * 12.4V * 1A * 7.5ms = 46.5mJ

the input current with my current probe set to 200mA/div.

so according to this, your latest scopeshot shows 560mA peak current.
The RMS current calculated by the scope is pretty useless because it is not calculated ONLY during the ON-pulse.

This screenshot was in a stopped state, i apparently forgot to save the same situation when running,

Does this overunity "backspike" energy recovery efficiency also occur when the rotor is stopped?

I remember the wattage calculation to be 35mW, so a difference of about 10mA between running (35mW) and stopped (44.78mW)

The average input power is pretty useless (just like the RMS current) unless it is measured only during the ON-pulse.
Does your scope have the ability to do measurements between cursors, or do you have to manipulate the display to show only the part of the waveform that you need to measure?

(Note that the current controller was set to 200mA/div, so the green current value and the wattage values need to be taken x 20,

Yes, 200mA/div is pretty clear.  That's how I got the 560mA peak current from your recent scopeshot.
It is also in the ballpark of the 880mA peak current that was shown on your previous scopeshot, so that confirms that this is not a x10 error or something silly like that.

Taking more samples did not change the Math mean value.

Your i*V multiplication by the Math channel is set up correctly. Only the calculation of the mean value is not narrowed down to the ON-period like it is supposed to be.

[gotoluc]

Thus the "backspike" recovery efficiency is E₂/E₁ = 12.4mJ / 120mJ = 10.3%

That kind of sucks

Thanks for taking the time to do what I can't do.

Luc

[verpies]

That kind of sucks
Thanks for taking the time to do what I can't do.

I expected 80% "backspike" energy recovery efficiency.

Both 10% and 125% are surprises to me