Muller Dynamo

[toranarod]

@Toranarod,
What is the purpose of the centertap? nice coils though.  Are you using a winding jig?  Would you show it if you are?

My home made winding setup. its a cut up old cordless drill with a counter glued to the side.

Oh the center tap ?
I am trying to get as much testing off the one setup as possible. I want to set the 300 turn setup and compare it to what I considered is better winding.

[JouleSeeker]

This is wrong!

You have then actually lost half of the energy  !!!

Very important error!

Stefan is correct:  CHARGE is conserved when a charged cap is connected to another cap, voltages are the same across the caps after connecting, and using E=1/2CV**2, you can determine that half the energy E is lost (in the wires, sparking, etc.)  (Take C1 = C2 for the two caps.)

[ElektroBaba]

Okay, thanks for clearing this hall sensor timing up.

I think that RomeroUK has really luck with his timing.
It is not better to use a simple light sensor (forked light barrier and a little hole, instead of hall sensor) and a simple electronic to control delay, width and height of the pulse?

[konehead]

hi Scratchrobot

your rig sounds really great! it really screams...very nice...

Looks good too except for that rotor eh...I bet you could cut the draw in half if you could get better airgap clearance between rotor magnets and cores liek 1.5mm....looks around 7 or 8mm or so right now judging from the video.

since its made of MDF (I assume rotor too) I would suggest using felt pen, and mark on rotor as it spins, a band on the outisde rim of rotor, so you can see exaclty where the rotor's "center-of-rim" or "centerline" is.....

then pop magnets out, and belt-sand that MDF rotor down and get out the wobble-ness...using the felt-pen marking as guide.
you dont have to get it perfect, but if you make the rotor thinner, letting the matgnet stick out a bit,  then you can play with each  magnet moving them up or down in rotor, until the magnets thenselves have the the perfect airgap and if rotor is little bit wobbly doesnt matter the magnets will be perfect....

anyways looks cool and goes great...i agree with bolt yours is going to be a looper judging by the way it goes already.

[conradelektro]

You see, the motoring coils are nr.1 and nr.6 The angle between those coils are 200 degrees. (360/9=40. There are 5 spaces between each coil, each space has 40 degrees, equals 200 degrees) On the rotor he got 8 little timing magnets each has a 45 degrees angle between them. We got 180 degrees between timing magnet nr.1 and timing magnet nr.5. So, when a timing magnet is in front of hall sensor that drives coil nr.1, the angle between the hall sensor that drives coil nr. 6 and the nearest timing magnet is 20 degrees. The conclusion: the motoring coils never fire in the same time due to the odd-even configuration between number of coils and number of magnets of Muller dynamo. Excuse my empirical math and my poor english. Hope it helps.

Hall sensors are not a very good switching means:

- As keykhin pointed out, one needs two hall sensors if one wants to switch two pairs of drive coils (e.g. coils pair 1 and 6), because the switching of coil pair 6 has to happen 20 degrees "later" than the switching of of coil pair 1.

- This clarifies when the drive coil pairs 1 and 6 have to be switched on. But when is a good moment to switch the drive coil pairs off?

Using hall sensors (or reed switches) allows to set "the switch on point" fairly accurately by placing them at a certain spot on the stator plate near each drive coil pair.

But the "the switch off point" is almost impossible to control, it depends on the size of the passing magnet and the sensitivity of the hall sensor (or reed switch). The release magnetic force is always different to the engage magnetic force in these sensors or switches.

Therefore I advocate switching by "angle" employing an encoder wheel (e.g. with 500 slits and a reset slit at 0 degrees). Unfortunately this implies microprocessor control.

The smaller the "switch on angle (= switch on time)" the less energy is needed for driving the motor. Of course, if the "switch on angle" is too small, the drive becomes inefficient. This needs careful calibration (which is not possible with a hall sensor or reed switch).

RomeroUK showed that it is somehow possible with hall sensors, but some "efficiency" might have been lost (because the drive coils are switched on too long). This might also explain why he needed a certain number of revolutions (and not a higher one). The switch on time of a reed sensor is only "just right" for a certain number of revolutions (rotor speed).

@ElektroBaba: the "switch on time" has to be defined by "angle" not by "time" if one wants to stay independent of rotor speed.

To make it even more complicated: due to delays in coil magnetization it might be necessary to even vary the "switch on angle" dependent on rotor speed. But as long as rotor speed is small in comparison to magnetization time a constant "switch on angle" might be sufficient.

Greetings, Conrad