Muller Dynamo

[mondrasek]

so probe 1 ground is connected as you show in your schematic. is your probe 2 ground connected to anything?

Probe 2 ground is not connected.

This is probably irrelevant, but whenever I tried to connect the second probe ground I would draw on the dynamo.  I guess this is because I was reading the current and voltage 180 degrees out of phase.  But that is what a simple ME would do, neh?

Let me know if you need any other testing or clarifications.

M.

[poynt99]

@pointy...erh, uhm, poynt99,

I was wondering if your simulation specialties could be requested for this latest RomeroUK Muller idea?  I know you heave electrical circuits mastered in simulations.  But can you do these moving EM interactions?

Which idea in particular? SPICE can handle non-linear cores and the flux changes going on in them. It is not a good simulator for pure magnetics and magnetic field interactions though.

I mean, particularly this case:  WTF is happening when we place a backing magnet?  The only theory that I have now is @Hoptoads explanation of the shift of the Bloch wall.  But one thing that is lacking from his explanation is the mathematical predictability of how much Gauss at what distance (and maybe offset relative to coil center?) makes what difference.

Do you have a link to Hoptoad's explanation?

A backing magnet pre-polarizes the coil cores, and this is the reason doing so increases the generated output. Without the pre-bias, very few of the core's domains are pre-aligned with each other. With core pre-bias, the core's domains are mostly aligned, and when influenced by a passing rotor magnet, they all rotate (or re-polarize in a different direction) in unison, thus generating a stronger flux change and hence stronger E-field build-up around the coils.

btw, there is no Bloch wall associated with a standard two-pole magnet like the types we are using here. There will be some Bloch walls in an un-biased core, but applying a pre-bias largely removes them.

Poynt99, do you have the facilities to investigate these variables?

With SPICE, not really.

Regards,
.99

[gyulasun]

Probe 2 ground is not connected.

This is probably irrelevant, but whenever I tried to connect the second probe ground I would draw on the dynamo. 
.....

Hi Mondrasek,

For most scopes with dual (or more) input channels the grounds normally are common grounds both at the body of the scope BNC sockets and at the crocodil grounds clips of the probes.  You can check this by a simple Ohm meter: you would find a short circuit between the ground clips.
So this is the reason why a certain (circuit) setup changes its operation  when you clip two probe grounds to two differrent points in it, you may simply short out a component.

In your corrected schematic (Reply #4150) you may wish to connect the ground crocodil clips of both probes to the right hand side of the 0.1 Ohm resistor (upper AC input point of the diode bridge) and probe 2 would connect to say the left hand side of the 0.1 Ohm and probe 1 would connect to the lower AC input point of the diode bridge.
This way when measuring points are chosen strategically the ground clips would not short out any two points or components where there is a voltage difference by the normal operation of a circuit.

Gyula

[slapper]

Probe 2 ground is not connected.

This is probably irrelevant, but whenever I tried to connect the second probe ground I would draw on the dynamo.  I guess this is because I was reading the current and voltage 180 degrees out of phase.  But that is what a simple ME would do, neh?

Let me know if you need any other testing or clarifications.

M.

i'm going to go on a few assumptions. you being simple ain't one of them. 

your inputs are non-isolated and your scope is performing a math function for the channel that displays voltage across the current sense resistor.
depending on what polarity the math is presented with (input 1 summed with inverted input 2; etc) is how it appears as if the scope was connected '180 degrees'.

if you tried to place probe 2 ground on the other side of your current sense resister while leaving probe 1 where it is in the schematic you would have, effectively shorted out the ac input side of your bridge and and probe 1 would see no voltage.

edit: (gyulasun beat me to it)

if these assumptions are correct we can move.

i wanted to highlight something you brought up earlier about core placement being very important.

the sim's that teslaalset did in his thread pretty much confirms this to me:
http://www.overunity.com/index.php?topic=10841.0

my testing confirms this as well. core location, and most likely the length, is important. and a pain in the a$$.

thanks.

nap

[mondrasek]

Which idea in particular? SPICE can handle non-linear cores and the flux changes going on in them. It is not a good simulator for pure magnetics and magnetic field interactions though.
Do you have a link to Hoptoad's explanation?

A backing magnet pre-polarizes the coil cores, and this is the reason doing so increases the generated output. Without the pre-bias, very few of the core's domains are pre-aligned with each other. With core pre-bias, the core's domains are mostly aligned, and when influenced by a passing rotor magnet, they all rotate (or re-polarize in a different direction) in unison, thus generating a stronger flux change and hence stronger E-field build-up around the coils.

btw, there is no Bloch wall associated with a standard two-pole magnet like the types we are using here. There will be some Bloch walls in an un-biased core, but applying a pre-bias largely removes them.
With SPICE, not really.

Regards,
.99

Excuse me Poynt99.  I presumed too much.  I believe the Bloch wall theory is on Page 5 here:  http://www.totallyamped.net/adams/

The "idea" that I was hoping you could simulate (or help facilitate) was why I saw an increase in V when a  PM was introduced at the correct, specific, distance, behind  the cores on my RomeroUK/Muller replication (w/rotor spinning).

Any "theories" are always welcome.

M.