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

[tinman]

Do you know how he has wound his coil and with what wire ?

I do now
But 10 ohm's is quite high,so i think i will drop that down a bit-or a lot,and go with a lower resistance across the coil. I will start with 2 ohms i think-less resistive losses that way,and see how it looks. It would only take 15 minutes to rewind a coil anyway,so might as well see if we get the same sort of results using a different resistance in the coil.

As i only have a 2 channel scope,the measurements will have to be done in pieces-this may make things a little harder--see how we go.


Brad

[MileHigh]

Let me give you same data to chew on:

According to the latest scopeshot, the energy expended by the power supply to bring L1's current from 0 to ~600mA is 17.745mJ, however according to the standard formula for energy stored in a coil E=½Li², the magnetic energy stored in a 37.8mH coil conducting ~600mA of current is 6.804mJ.
That is a big difference between what was consumed and what was stored in L1

The missing  ~11mJ  is easily explained by the mechanical energy gain of the rotor as well as hysteresis and resistive losses, but maybe this calculated energy imbalance will point you to a better explanation, what is going on.

I am not following the thread closely but I will assume that the timing is still coming from Itsu's signal generator and the rotor is slaving to that.  I agree that the missing energy is most likely explained by the energy gain of the rotor.

To me that suggests an important test and I hope Itsu would consider doing it (if not already done):  Just look at the coil traces when there is no rotor spinning and compare them with the case with the rotor spinnng, exactly the same pulse timing.  I believe a lot can be learned from that and it will open up people's minds to a series of issues that many readers are probably not thinking about or aware of.

I am a little confused though because I thought that Itsu's measurements are showing more energy in the back spike than energy expended to energize the coil.  Am I missing something?  As a result I am not sure what the better explanation is about since things seem to have returned to normal?  Sorry, I don't have the energy to read the entire thread.

MileHigh

[tinman]

I am not following the thread closely but I will assume that the timing is still coming from Itsu's signal generator and the rotor is slaving to that.  I agree that the missing energy is most likely explained by the energy gain of the rotor.

To me that suggests an important test and I hope Itsu would consider doing it (if not already done):  Just look at the coil traces when there is no rotor spinning and compare them with the case with the rotor spinnng, exactly the same pulse timing.  I believe a lot can be learned from that and it will open up people's minds to a series of issues that many readers are probably not thinking about or aware of.



MileHigh

I am a little confused though because I thought that Itsu's measurements are showing more energy in the back spike than energy expended to energize the coil.  Am I missing something?  As a result I am not sure what the better explanation is about since things seem to have returned to normal?  Sorry, I don't have the energy to read the entire thread.

Yes MH,Itsu's measurements are showing more energy captured in the back spike by the cap.
Verpies was showing what should be stored as magnetic energy in the coil,and what was actually consumed by the coil
Quote- :the energy expended by the power supply to bring L1's current from 0 to ~600mA is 17.745mJ-->however according to the standard formula for energy stored in a coil E=½Li2, the magnetic energy stored in a 37.8mH coil conducting ~600mA of current is 6.804mJ.


The recovered energy MH in C2 was-  E2 = ½ * 10μF * (99V)2 = ½ * 0.000010 * 9801 = 0.049005J = 49.005mJ

So 17.745mJ in,coil should only have 6.804mJ of stored magnetic energy,but returned 49.005mJ to C2

Brad

[verpies]

I am not following the thread closely but I will assume that the timing is still coming from Itsu's signal generator and the rotor is slaving to that. 

Yes AFAIK, except he is using a 555 astable to avoid grounding issues.  I think he is working on triggering from the rotor's angular position sensor for his next version.

To me that suggests an important test and I hope Itsu would consider doing it (if not already done)

Already done, albeit not with the latest version.

I am a little confused though because I thought that Itsu's measurements are showing more energy in the back spike than energy expended to energize the coil.  Am I missing something?

You are not.

According to the most recent scopeshot:

A)   17.745mJ  Energy expended by the power supply during one ON-pulse, to bring coil's current from 0 to ~600mA,
B)     6.804mJ  Calculated energy stored in the 37.8mH coil at the end of the ON-pulse,
C)   49.005mJ  Recovered energy in C2 after one ON-pulse.

The energy lost to the rotor's motion as well as the energy loss due to hysteresis are unknown.

The energy loss in the resistance could've been easily calculated from the RMS value of the current, but I prefer to have the peak current displayed on the scope and Itsu's scope can display only one statistic per channel ...I think.

I divided C by A in order to get the overall efficiency of 276%.    Do you realize what would've happened if I had divided C by B ? 

[gyulasun]

....

I tried that several times with different pulse motors and different drive circuits. I never could prove that input power to the motor was less if one tries to feed back power from the spikes (from the back EMF of the drive coil).

I think that the power in the spikes is forever lost.
...

Hi Conrad,

Would like to return to your above post, and I am aware of verpies's answer to it here:
http://overunity.com/16167/sharing-ideas-on-how-to-make-a-more-efficent-motor-using-flyback-moderated/msg469191/#msg469191  with which I agree.

I have drawn a possible schematic (attached below) where you can see two coils in close magnetic coupling (simplest way to construct them is to use bifilar windings). The second coil is used to feed the voltage spike back to the positive supply rail via the flyback diode. There is a capacitor across the supply rail and the flyback energy can charge it up whenever the magnetic field collapses just after the switch is off.
So from the second ON pulse onward, the main coil in series with the switch is fed from a higher supply voltage level what the input supply would normally establishes, at least for a certain part during the ON time till the capacitor discharges to as low as the input supply voltage level if that is allowed by the timing.
The second diode put in series with the positive input supply rail automatically isolates the capacitor from the supply voltage whenever the voltage level across the capacitor is higher than the input supply voltage.
The second diode certainly causes a forward voltage drop in the input supply, low barrier Schottky or a germanium diode is to be used to minimize this loss.

Low duty cycle i.e. short ON time is needed to reduce coil losses, this needs the L/R time constant of the main coil to be considered.

Other circuit solutions are certainly possible for feeding the flyback pulse back to the input supply rail, to be able to increase overal performance, hence increasing efficiency.

Greetings
Gyula