"Invisible fence" for dogs

Got any references to the theory? Are you sure "your experiment" was responding to H field ONLY?

no I don't accept that.

the magnetic field inside a loop is constant in the plane of the loop

Don't worry, I won't say it a third time.

Mark

Not silly to me... I've still got grounding and shielding confusions.

I took two air coils, aligned along z-axis, stimulated one and picked up with the other. I stuck a piece of 1/16" Al between them. No change for freq < ~100 Hz. I started to see attenuation at ~1kHz. (With also a bit of phase shift ?)

Oh, (head slap), I was picturing the currents wrong! (Thanks) So if one connected each end with a loop of wire...?

George H.

In what direction do you think that current runs?

Jeroen Belleman

But this is not a solenoid! A solenoid is long compared to the encircled surface. This is one single-turn encircling a relatively huge surface. A dog fence. Not at all the same geometry.

My probe is a small coil with just a few turns of thick wire between the pins of a coaxial connector, not a geometry very receptive to E-fields. It's easy to check for that, too, just by rotating the coil 180 degrees so the H-field changes sign, while the E-field does not. Should be clearly visible if that was the problem. It's not.

The theory is simply Ampere's law plus superposition.

Jeroen Belleman

Twas I that mixed up the current direction. Redrawing it, the tube and loop is a one turn transformer.. (which Mike k. mentioned up thread) And the transformer has to have the same current... which turns into another piece of the dog fence. (a fun puzzle.) Thanks again for the experiment.

George H.

Did you read this part...... the more surprising thing about the magnetic f ield inside a solenoid is not that it's uniform along the length, but that it's uniform in the perpendicular directions -- that is, that the field doe sn't depend on whether you're close to the axis or far from it (as long as you're inside it). It'd be easy to imagine the field would either drop off or get stronger as you move perpendicular to the axis, but it doesn't (agai n, for a long solenoid when you're not near the ends).

Then the pipe and the loop of wire would constitute a secondary winding transformer-coupled to the primary loop. By Lenz's law, it'd attenuate the B-field around the pipe-enclosed primary wiring, and the secondary loop might carry enough current to make the fence effect.

Too hot in the garage.

We've never had ticks as badly as we had them in NY (and at my grandparents in the MI UP). Actually, the bugs in the South have been minimal. We did have a scorpion in the living room last summer. One of the cats was playing with it.

0.5mm from the inner surface of the pipe, it is antiparallel to the enclosed wire current. 1mm from the inner surface of the pipe, it is parallel to the enclosed wire current. 2mm from the inner surface of the pipe, there is negligible current. The outer diameter of the pipe is 3mm from the inner surface.

I already answered that, it will stuff everything up.

it will reduce the magnetic field round the whole perimiter and concentrating the electric field at the feed point and at the locaton of the beads.

--
  \_(?)_

the pipe has much larger cross section than the wire, the resistance added is probably insignificant.

--
  \_(?)_

That's an over-simplification, if you get close enough to wire you'll see the field strength following a 1/r curve but once you get 1 quarter of the way across the yard the field looks pretty reasonably constant.

If you cycle on the streets when there's little traffic you'll see that a bicycle doesn't have enough metal to trigger the buried metal detector loop that drives the traffic lights if parked in the centre of the loop, but put one rim over the seam where they cut the loop in and the light will soon go green.

clearly the field is stronger next to the wire.

nah it wouldn't.

--
  \_(?)_

OK, you want to talk skin depth. At least you have the direction right. Do I read you correctly? Do you state there is no current on the outer surface because it's several skin-depths away from the inner surface?

At my 1MHz frequency, the current essentially flows over the surface everywhere, anti-parallel inside the tube, indeed, but parallel on the outside surface. There's almost nothing in the bulk metal.

Jeroen Belleman

perimiter of the loop?

the plane of the loop.

f

op would active the collar.

OK, you are probably right.

The field near of an infinte straight wire varies as is 1/r, no issue there .

If we start with a RECTANGULAR loop say 100 feet on a side, when you get cl ose enough to one wire, it approximates to an infinte straight wire. So I agree in that case.

So then I thought perhaps a CIRCULAR loop might be different. But no, as you get closer and closer to the wire of a circular loop, the cu rvature appears to reduce and again we approach the infinte straight wire c ase.

So I guess you are correct, the field may be relatively constant inside the loop away from any wire, but as you get "close enough" to any wire, it __ must__ approach the infinite straight wire case which is of course 1/r.

Thanks for maintaining a civil discussion, a rarity on usenet these days.

Mark

Wow - what a long discussion. I couldn't follow most of it and since there was so much disagreement, I wouldn't have known what to believe anyhow.

I tried the iron pipe thing - I used a 1/2" Sch 40 pipe about 4' long. I could not detect any difference in the receiver's detection.

My solution was to run the wire in the basement, low enough to not trigger when passed over (3=4' min). Otherwise, I probably would have used Dan's idea of paralleling 2 more wires several feet apart.

Thanks for all the replies, Bob

[about wire carrying signal, inside a buried pipe section]

Are you familiar at all with skin depth? Aluminum, at 1 MHz, has a skin depth of 82 um, and iron pipe would be an order of magnitude less, because it's MAGNETIC. That current distribution you describe is of a pipe magnetized in the circumferential direction, but you can't magnetize iron that fast: that's why megahertz inductors have ferrite or air cores.

Sigh. *Read* what I wrote.

Jeroen Belleman

I was thinking the absence of B-field in the pipe wall meant no inner surface currrent returning on the outside of the pipe; I was wrong, your picture (wire and inner-surface currents in opposition, outer-surface current parallel to and near-equal to wire current) is the representation of circulating current that PREVENTS magnetization in the circumferential direction of the steel pipe. I was looking only at the pipe circulating currents, forgot the wire current's induction was also present.

Text-only is a hard medium for communication. I kinda miss the long corridors lined with chalkboards, so anyone wandering the old physics building could doodle without delay.

I had decided to drop out of this discussion, or what was left of it, but to conclude, yes, I agree.

Well, that's Usenet.

Jeroen Belleman