Muller Dynamo

[Dave45]

@Bruce_TPU & RomeroUK, haven't built anything yet but I think I get where you're going. Its just like what Naudin did with the miniromag gen? http://jnaudin.free.fr/html/mromexp.htm except you were using two coils? Quote "A simple diode has been used to short the back EMF part, and you can notice that the rotor speed remains constant...." So to bypass lenz, we don't use/cancel the induced EMF and use the energy/spike from the collapsing field?

I think folks need to take a second look at this

[chrisC]

Hello Everyone,

I have been quietly watching the progression of this thread for several days now, and though I have been working mostly in the background to understand the basics of this stuff, I thought I would finally chime in and contribute something.

I think that you all should really pay more attention to the posts made by Bolt. He is very much a senior member on this forum and has contributed many, many insightful ideas that should not be overlooked. In particular, the post he made here was particularly instructive to my understanding:

http://www.overunity.com/index.php?topic=3842.msg290450#msg290450

The above post is simply a reiteration of what Bolt has been trying to speak to everyone since the beginning of this thread, and it took me a couple of read-throughs to really understand the essence of what he was trying to explain. But to help me really intuitively understand what was happening, I took some time and drew up a few quick simulations in the Java circuit simulator to observe the effects of tuning to series resonance on a simple 1 to 1 transformer. Recall that the primary "effect" of the coil acceleration is due to the 90-degree phase shift of voltage to current in the generator coils.

The first attached picture at the bottom of this post is a screenshot of the circuit that I ultimately ended up with.

As I continued to play with the idea, my thought was to model the circuit using the real inductance values of the pulse motor I have (which are 1.1 mH and 0.5 Ohms). So, I modeled the transformer as a 1:1 winding ratio with a inductive coupling coefficient of 0.3 (so basically the coils were loosely coupled magnetically). This was meant to simulate the magnets moving past the coil and electrically, the behavior is close enough to simulate with reasonable accuracy.

The 400 Hz input frequency was arrived at by calculating the speed of the motor to be 6000 RPMs. At this speed, the tuning capacitances would be low enough to realistically deal with. After setting these values, I calculated the necessary capacitance to put the generator coil into resonance (which was 143.9 uF). I also added the same value to the input side in a parallel resonance circuit to reduce the input power requirements as well (not required since the input represents a magnet rather than transformer).

The two 100uF capacitors may not necessarily need to be there but I found that the output current and voltage waveforms on the coil looked cleaner with them present. My assumption for the time being is that it somewhat decouples the output circuit from the resonant tank circuit of the coil. As for the DC Bridge rectifier portion of the circuit, I just arbitrarily choose a value for the DC smoothing cap and load resistor.

What I found was that when the output coil was properly tuned to resonance, the current waveform coming out of the coil did lag the voltage waveform by 90 degrees. But when the circuit got detuned, the voltage and current waveforms would be aligned with eachother. Factors that initially affected this were the resistance of the coil (represented by the 500mOhm resistance), and the load resistance on the bridge rectifier.

After playing with the model more, I finally found that the load resistance did not change the phasing of the coil as long as the 100uF caps were in place, Without them the waveform became heavily distorted which made it hard to make an accurate phase measurement.

However, the most wild thing about this particular circuit was the input and output values that were observed. I set the load resistance to 20Ohms with the input voltage being 300V (assumed to be the open-circuit output voltage of my pulse motor's coils at 6000 RPMs). With these settings, the input power was only about 20-30W peak while the output power across the 20 Ohm load was close to 1.7 kW. Of course, I was quite skeptical of these results and attempted to reproduce the schematic in Multisim to confirm the results that I was seeing. I took a screenshot of the setup (Attached below).

Using realistic circuit models for the bridge rectifier, I found that the output power was far lower than what the Java simulater predicted, however, I was very pleased to find that the phase behavior of the coil with and without resonance, and with light and heavy loads agreed 100% with the results I obtained in the Java simulator.

So now armed with these confirmations, the question to ask is why the coils would cause the generator to accelerate under load rather than decelerate. The simple answer is because the delay in current production also causes a delay in the magnetic field production. Since under normal circumstances, the induced voltage and current are in phase with eachother, the magnet experiences drag as soon as its movement induced an EMF in the coil. However in the resonant state, the magnetic field would be delayed by up to 90 degrees from the initial EMF.

For most coil geometries (particularly, a simple solenoid coil), the induced voltage positive and negative peaks occur when the magnet’s center is directly over the edge of the coil. This means that in the resonant state, the current would not even begin to rise until the magnet is almost under the coil or at top dead center (TDC). At this point, that means that any repulsion force generated by the coil would either have a reduced effect on the magnet or actually accelerate the magnet out from under the coil if the delay was far enough. This explains why high impedance coils naturally posses this quality â€" simply because the impedance of the coil added a sufficient delay to the generation of the magnetic field so as to give the moving magnet enough time to move under and away from the coil before experiencing significant drag.

- Jason O

EDIT: Also, forgot to mention this but for those who would like to get up to speed on the basics of AC rective power and VARS, check out this video here. The author does a pretty good job of explaining things and has a bunch of other nice videos on his YouTube channel explaining other topics with AC power:

http://www.youtube.com/watch?v=g0S-XV-BiUA&feature=related

@jason

Thanks for the simulation results. I totally agree with you about the key to possible O.U is where the current and voltage are tuned to 90 degree out of phase as Bolt and Romero both pointed to. Romero even stated that with one driver and one pick up, even tuned properly will not enough to generate enough juice to loop properly. I believe once people get to the part where they can tune their different builds properly, some may be able to replicate Romero's device but to to it consistently, the physics and mechanics must be properly understood. Great job and thanks for your confirming simulations. Looking forward to more sharing from you.

cheers
chrisC

[e2matrix]

Hi Remero,

I see what it is now ... it's your magnets!  When spinning it looked like a aluminum cylinder around the disk platter, I was trying to understand what it was (circle in red) and where your magnets were on the platter.

I guess HDD has 2 or more platters and the magnets rest on both and are held with tape and maybe a little Super Glue.

Great idea ... I think that would be small enough to experiment in my RV home

Thanks for the reply and this great idea

Luc

ADDED BTW could you tell me the size of those bar magnets and the diameter of the HDD platter, as it all fit very nicely!  I'll see if I can find the same sizes. Thank you

gotoluc,  having taken apart a lot of hard drives I'd say that is probably a 3½" hard drive.  Almost all hard drives other than real dinosaur's are 3½" hard drives and that is the platter size you see when you take them apart.  It's possible it could be a 5¼" (your very old 20 Megabyte to 140 Megabyte size hard drives from the late 80's to early 90's) platter HD but without much point of reference for size I'd still put my money on a 3½" based on platter color and commonality.  Laptop drives are generally 2½" so I think we can rule that out.  I've got a hard drive with about ten 14" platters but that's about as big and rare as you'll find so I doubt that's what he's got either    HDD trivia lesson done.     

[gotoluc]

Thanks e2matrix,

so 3.5" is = about 11" or 280mm in circumference

11 inches / 30 magnets = 0.367 inch or 9.33mm wide. So 28 of 10mm wide magnets should fit well.

Thanks for sharing

Luc

[penno64]

Isn't it marvelous how a mistake can lead to discovery!

Penno