Thane Heins Perepiteia Replications

[minde4000]

Here is same E coil with bifilar 27 awg 43 ohm outter coil. There are 10 layers. However this outter coil acts as a brake at any time. Voltage does triple across the load when inner is shorted but I think I didnt wind enough ohms. Need to wind an extension I guess to some 60-70 ohms and see what happens then.

EDIT: voltage generation open circuit is like 1:10 less value being outter

@Baroutologos

Sorry man I will not measure anything right now. Even tho inner is accelerating with power production it does not overcome cogging torque all by itself at any time so there is too much power consumption by ryobi. If outter will work out with extra winds I will measure then but till then not too much "perpetia" in it...   I have a feeling I cores large coils with low awg will give a better perfomance overall as is was said by T.

Regards Minde   

[icanbeatbob]

Minde.
Your coils are awsome. I envy your work. Good luck on the replications.

Brad

[baroutologos]

@Minde,

You winding skills leave me speechless... You are gifted Minde on that. You are a natural... I will never post my windings!

And yes, as you say, everything depends on:
how strong your magnets are
- mine are 40 libs strentgh doubled stacked (25mm x 10mm N40)

how much air gap
- i operate with two coils on two magnets simultaneously at some 1 cm distance. My SKF 20mm inner diameter bearings make tremendous noise and vibration if i dare to lessen the gap. Need odd-even configuration here. Grinder's bearings are specially made for applying some short of unbalanced force on the spinning grinder stone and still being silent.

how strong magnetic flux is put through your cores
- limitations exist though that can lead to core saturation.
I find out that by physically attaching a coil to a laminated surface for closing the magnetic path (or backing toroid i think ) is not as good as having solid E-core since it leaves some tiny airgap anyway.

And of course... coil's geometry and composition.
-AWG, number of turns, winding skills :p etc.
I have reached the conclusion that above a certain layer's thickness, more layers are burden. You can better go with a longer core to have wire turns as close as possible to core.
You can achieve same results by incresing core diameter but not too good though, since crossection increases exponentialy to radius in contrast to surface which increases proportionally. (we are after surface since layers are there) meaning more mass, more losses, etc etc
...................

Ok, all these are already known i suppose. I felt a small repetition does not hurt anybody
I look forward this weekend to experiment with output generation and how can output can be efficiently stored. We ae working, if done corrected with the biffilar setup at kilovolt range. WTF? Using step down transformers for that range are of any good? What about at using many coils... many stepdown transformers? Geez... must be another way.
(Mr Thane would not respond on cap filling )

I must admit that my inner desire is to make Perepiteia to go OU and ultimately become a self runner. It's not a shame to admit that i suppose, after all we are all seeking that with one or another way.

Regards,
Baroutologos

[CRANKYpants]

Here is same E coil with bifilar 27 awg 43 ohm outter coil. There are 10 layers. However this outter coil acts as a brake at any time. Voltage does triple across the load when inner is shorted but I think I didnt wind enough ohms. Need to wind an extension I guess to some 60-70 ohms and see what happens then.

EDIT: voltage generation open circuit is like 1:10 less value being outter

I have a feeling I cores large coils with low awg will give a better perfomance overall as is was said by T.

Regards Minde

YOU ARE RIGHT - AND I THINK YOU SHOULD STICK IT OUT WITH YOU E COILS FOR A WHILE MINDE - I AM SURE IT'S NOT EASY FOR YOU BUT IT IS INSPIRING FOR EVERYONE.

YES - WHAT YOU NEED TO DO NOW IF POSSIBLE? IS TO RECORD YOUR PERFORMANCE DATA,
THEN ADD ABOUT 20 MORE OHMS AND THEN RE-TEST.

WHAT YOU OUGHT TO FIND IS SLIGHTLY LESS DECELERATION.

THEN ADD MORE TURNS (AGAIN ABOUT 20 OHMS) - AND RE-TEST.

KEEP DOING THIS UNTIL YOU FIND THE NEUTRAL ZONE AND THEN BEYOND THAT THE ACCELERATING ZONE - AS IN THIS GOOD FRIDAY VIDEO:
http://www.youtube.com/watch?v=Lp2x6Vtu9Kg&feature=channel_page

THE IMPORTANT THING IS THAT YOU HAVE ESTABLISHED THAT YOU CAN COLLECT INNER HV COIL FLUX AND ADD IT TO THE OUTER HV FLUX. THIS WOULD BE A GOOD GENERATOR IN AN ELECTRIC VEHICLE BECAUSE NOW YOU HAVE SHOWN INNER COIL REGENERATIVE ACCELERATION AND OUTER COIL REGENERATIVE BRAKING.

GOOD JOB! 

T

[wattsup]

What I understand so far is that one of the main variables between the shown builds is the space between magnets of the rotor. If the space is greater, each magnet pass will cause a greater whipping effect off the coil since the zero and TDC will be at greater extremes. If the magnets are closer, the cycle changes will be less extreme since the next magnet is already on the core before the previous one leaves the core. This to me spells a rotor having wider magnet spacing wants to impart more amperage but is stuck in voltage mode because of the coil design.

I suspect the coil, in the case of wider magnet spacing, should be multiple wires all wound together, like 4 to 10 or more wires of 18 to 22 awg all wound together over the core, then connected in parallel. This will lower the output voltage but increase the amperage making it easier to then send to any electrical device. As a shorted coil, this may also provide a better and quicker and stronger repulse to cause a better acceleration effect.

The bifilar may be good at the higher frequency but only when the core can handle the core to coil frequency like Thanes core material.

But if you use a regular transformer core that is rated to be used at 60hz, you may have trouble.

Example:
At 2400 rpm with 12 magnets, if each two magnet passes (one north then one south) is one cycle this equals 14,400 cycles per minute or 240 hertz. If you used a 60hz compatible core you would need to have a winding with 4 parallel wires. Your wheel will turn at 2400 rpm, but the coil will cut the core transfer into a lower frequency but bigger amps. If you used 8 wires, you could then bifilar them back to 4. If you used 12 wires, you could trifilar them back to 4 also.

I think that's why the bifilar coil is heating up. It does not know what to do with the high frequency since the core cannot react fast enough to the coil flyback that makes the repulse acceleration, it is instead getting stuck in the coil and creating heat. A simple test would be to just connect the now bifilar coil as a simple dual parallel coil. This should double the amperage, half the voltage while halfing the flyback speed and hopefully reduce some of the heat. Wow this adds to the overall dynamics of the many potential effects.

Once a rotor is made, the number of magnets and spacing cannot be changed if the rotor design does not permit sliding the magnets to add more. Only the rpm can be varied but you need a minimum to start any reasonable transfer. But you can play with the number of wires in your coil to decide how to translate the rotors' eventual output that should be based on the magnets rpm but also on its spacing.

Just look at a regular car alternator stator. Why do they use three phased coiling of good sized wire to then rectify it to 12vdc but with 100's of amps with final result of 1200 watts. How many turns are in an alternator stator per phase? Not many. Hmmmmmm. Yes the available torque is not the same but the transfer method should tell us something.

So it looks like there is definitely more things to consider before I make my final rotor design, magnets with fixed or variable spacings, rpm, available torque, core material, wire size and winding methods. It is one major balancing act. I think the rotor/magnets should be made first, then matched to a motor, then driven under various simulated drag conditions to find the most efficient input/rpm versus output. Then take that rpm, add an expected rpm increase under the shorted coil method, and with that final rpm, build the core/coils according to the desired acceleration and/or output.

I'm still waiting for a reply from the C core manufacturer. I guess my realistic build timing will be by mid to end of August depending on what the C core manufacturer says.

I would also like to give my thanks to all you guys who are sharing here since all this information is critical for me to prepare my build.