The TPU uncovered? (A PROBABLE technique.)

[poynt99]

@all

Here is the coil current flow pattern. Remember, the current is flowing from the inside of the inner coil, to the outside, while the current is flowing from the outside of the outer coil towards the inside with the current direction arrows as indicated.

By definition they are hooked in parallel, but in current flow they are parallel mirror opposites. That is why a regular schematic or simulation program can be worthless, unless it can differentiate pancake coil connections and action. If spice can, that would be great.

Paul Andrulis

it's simpler than this. pancake coils (not necessary), cw/ccw windings (not necessary), schematics not able to represent what's going on? (not true) etc.---try to keep it simple,  overcomplicating it makes it less clear.

[pauldude000]

@Duff

I have been puzzling out your setup and I have one thing to say. I am lousy at explanation.

Dang, maybe this is too complicated of a setup. I made it opposing in all factors. Direction of rotation of current, spatial directional current flow, and generated magnetic field alignment. It is designed for all factors to hit an electrical "brick wall". The harder this is hit, the harder the wall.

You have two to many coils in your setup. There are no resistors. The necessary resistance is provided by each coils self-inductance and mutual opposing fields. The coils are not center tapped. (I bet you thought 3 and 4 were nodes. My fault not yours.)

It is an interesting concept though. the current flow would be strange, along with the magnetic fields..... kewl. A simple schematic of two coils in parallel is more accurate though.

By the way, if you build, try the magnet exeriment I suggested to poynt earlier.

Paul Andrulis

[pauldude000]

@poynt99

Hmmm......... Coil type doesn't matter....... Winding direction doesn't matter.......


Basically you are stating is is an incorrect usage of the scope connections? (It is possible, I am not an expert.) 

Or  are you saying a basic principle? (Equally possible, as I tend to get lost in field examination sometimes.)

I have to disagree on the schematics statement though. Unless your schematics show the winding relationship and style of coil, then no conception of current flow/vs/magnetic field alignment can be attained without extra in detail winding explanation also provided.

Paul Andrulis

[poynt99]

@poynt99

Hmmm......... Coil type doesn't matter....... Winding direction doesn't matter.......


Basically you are stating is is an incorrect usage of the scope connections? (It is possible, I am not an expert.) 

Or  are you saying a basic principle? (Equally possible, as I tend to get lost in field examination sometimes.)

I have to disagree on the schematics statement though. Unless your schematics show the winding relationship and style of coil, then no conception of current flow/vs/magnetic field alignment can be attained without extra in detail winding explanation also provided.

Paul Andrulis

Paul,

see my diagram above. i've marked the poles of each coil so you know by looking that the currents are running in opposite directions for the two. that's how it is done.

the type of transformer is not critical, but some will work better than others. the method you used is not ideal. even though it looks like your two concentric coils worked, a better way would have been to wind a single pancake coil, but as bifilar (both wires wound in the same direction). then hook them up as i've shown in opposition. the problem with yours is the two coils have unequal inductances, and the coupling is poor and uneven between them.

your scope connection is ok. your basic principle is fine too. you are canceling the fields, which is the goal. winding the two coils as i described will give better canceling as they are both occupying the same space (tighter and more consistent coupling), and they both have very close to the same inductance (will react identically to the input pulse).

you could have used a basic bifilar solenoid coil (a stretched out pancake coil), or a bifilar toroid too. results should be similar.

[duff]

@Duff

I have been puzzling out your setup and I have one thing to say. I am lousy at explanation.

Dang, maybe this is too complicated of a setup. I made it opposing in all factors. Direction of rotation of current, spatial directional current flow, and generated magnetic field alignment. It is designed for all factors to hit an electrical "brick wall". The harder this is hit, the harder the wall.

You have two to many coils in your setup. There are no resistors. The necessary resistance is provided by each coils self-inductance and mutual opposing fields. The coils are not center tapped. (I bet you thought 3 and 4 were nodes. My fault not yours.)

It is an interesting concept though. the current flow would be strange, along with the magnetic fields..... kewl. A simple schematic of two coils in parallel is more accurate though.

By the way, if you build, try the magnet exeriment I suggested to poynt earlier.

Paul Andrulis

@paul

The resistors are a requirement of the simulation program when making the connection between a primary & secondary.

I considered L1 the primary and L2 the secondary ( note the value of the resistor 1 x 10¹² ).

Why?

That's one of the ways that spice's output made sense.

I tried modeling it exactly as you posted it but the model's output was not even close to what you were getting (same as input or no signal).



@poynt99

duff thanks for the modeling. what do you conclude from your particular (slightly odd) circuit and its results? is your circuit correct?

The circuit I posted was ONE of the ways I tried modeling it and is somewhat of a challange to get to work.

As to being correct, well due to the odd way the circuit is wired I can only guess.
I have not drawn any conclusion at this point - still thinking about other ways it could be modeled.

why are you measuring between node 1 and 2? is anyone paying attention to this...pick up an error maybe?

Oddly, that is the point which displayed the output wave form that Paul posted. The problem is the model's output is in the microvolts...


@paul

I'd be interested in seeing your output between node 1 & 2



-Duff