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

[shylo]

Is this the same?
I'm collecting the back spike in caps, I'm using the banks of caps to drive the rotor,but when one bank is driving the other is outputting, but the output adds to the drive.
Sorry I haven't watched the video's ,can't afford it.
Lots to do.
artv

[gotoluc]

BTW: How do you like positioning the driving pulse electronically, without having to physically move the opto?

I love it!... verrry accurate and saves so much time. No one knows how much time it takes to change the width of the reflective strips to adjust the on time. On the fly pulse width is a must option. Timing is also nice to have and again much more accurate.

These are nice looking coils.  Did you glue the windings together?

Yes, the new drive coil was super glued after it was wound and its side plates (to support the wire while winding) carefully removed, then more super glue added on the sides to keep the wire from falling out the side. Kind of a trick build.

What are their inductances and resistances ?

The drive coils is 0.55 Ohm with 3.76mH
The generator coil is 8.1 Ohms with 150mH

This motor has three outputs:
1) The 100Ω resistor across the "generating coil"
2) "flyback" recovery cap and "draining pot"
3) Mechanical

The average power dissipated #1 is the easiest to calculate as P=V_RMS²/R which calculates to: P = (7.12V)²/100Ω = 0.507W or  507mW.

The average "flyback" energy #2 captured in 1 cycle is the peak voltage across the recovery capacitor squared, times capacitance and divided by two or E=½CV² , which calculates to: E=½*34.2μF*(10V)² = ½ * 0.0000342F *100V² = 0.00171J = 1.71mJ
But the motor runs at about 65Hz under load (pls verify), so we have 65 * 1.71mJ = 111.15mJ per second or 111.15mW.
The average "flyback" power is actually more than 111mW, because that "draining pot" keeps draining the 34.2μF cap even while this cap is being charged. (no periodic discharge circuit like Itsu's). 
...and if the "draining pot" is wirewound and inductive, that it forms an LCR circuit with the 34.2μF cap creating an interesting problem...

Agreed,  the generator coil is delivering 507mW plus its internal resistance which you can determine with the now known wire resistance of 8.1 Ohms

I checked the frequency and it's 88Hz (9 samples in window @10ms), so we have 88 * 1.71mJ =150.48mJ per second or 150mW on flyback recovery.
I agree, we can't calculate all the Joules from the cap the way I have it set up. I can live with that

The mechanical output #3 remains a mystery unless you are willing to calculate the bearing and windage losses ...or play with prony brakes.

I wasn't interested of the mechanical output of the rotor. However, an idling motors rotor has zero torque.
What I was interested in knowing, is the motor electrical to mechanical Watts efficiency.
So in my test video I used a known load (a generator coil with known resistor) 
I applied the generator coil to load the motors magnet rotor to give me an idea of the motors efficiency.
As we know a generator coil is never 100% efficient because of wire resistance losses (in my case 8 Ohms) and core losses.
So a small part of the power is lost in wire resistance and part is delivered to the load which we calculated to be 507mW across the 100 Ohm load resistor.
I'll let you calculate the power dissipated in the wire resistance.

Here is the problem with my test results. The generator coil is delivering 507mW plus some extra mW caused by its wire resistance.
How can we explain the motors power input only increasing by 350mW? ...  like I said and I've even read MH say the same thing, an idling motor has zero torque. So if there is no continues power storage in an idling motors rotor, then when we apply a load to the rotor, it should consume more power from the input and normally a little more power then the power the load is delivering. 
However, it's not the case, we have a shortage of over 150mW!... why?

Can we have the schematic of the power pulsing circuit ?  with the power transistor, flyback diode and all...
Also, it would be nice if you'd care to scope the signals going to these multimeters by using AC channel coupling on your scope.  Watch for a groundloop!

I'll see what I can do to provide a schematic. You know how much I hate doing schematics
I'm not sure why or how to scope the signals going to my multimeters using AC channel coupling?
I'm aware of the groundloop problems

If you keep mentioning incremental input power in comparison to absolute output power, then a cerberus will descend from a high mountain and chew you out. From what height?  ...about a mile, I guess.

Yes LOL

P.S.
The current waveforms are not easily explainable on your scopeshots.  I can smell a "back EMF" discussion brewing.

Let me know if you want me to post some zoomed in scope shots.
As long as discussion are kept civil, respectful and productive they can be entertained in this topic.

Kind regards

Luc

[gotoluc]

Is this the same?
I'm collecting the back spike in caps, I'm using the banks of caps to drive the rotor,but when one bank is driving the other is outputting, but the output adds to the drive.
Sorry I haven't watched the video's ,can't afford it.
Lots to do.
artv

Sounds similar.

Do you have a laptop?

Luc

[itsu]

Hi Itsu

Not an easy result to achieve!... unless coil size and copper mass does not matter?

try the attached coil calculator program.
Results are air core.
I've used it for years, with accurate results.

Let me know if you have any questions.

Luc

Thanks Luc,


i had some stuff laying around, so i used that for building my new coil.

I see you are going on very well to (very nice coils), and it seems we are using almost the same things to build the coil.

I build my core from metglas tape also, but did not chop it up, just rolled about 8m of this tape onto a fiberglass rod
to create a metglas core of 4 cm long, 18mm thick to fit a former, see video: https://www.youtube.com/watch?v=YSB-fj1PH9w&feature=youtu.be

The new coil measures 0.8 Ohm @ 4.5mH and is using 1mm diameter bonded magnet wire, bifilar wound.

Due to this low resistance i had to lower the ontime (duty cycle) of the pulse, but with verpies his sequencer that was no problem.
It now runs at 40Hz (won't go any higher) @ 4% duty cycle (was 20%, but won't go any lower then 4%).

Its amazing what can be done with a few cents worth of chips/components and some time as this "verpies sequencer" gives you full control over the pulses used.

I filled in the spreadsheet below with this new data and we see a big improvement on the "input recovered in C2"
At the same time the "unaccounted Energy" part is reduced strongly compared with the earlier coil/setup.

The screenshot show some data collection points to fill in the spreadsheet which is added to.


Regards Itsu

[gotoluc]

Excellent drive coil Itsu!... which gives excellent test results as well.

Are you going to take it further by trying 2mm wire?  as I see your spool has room for twice the amount of wire

Thanks for making a video of your coil build and posting all the test data results.

Kind regards

Luc