[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Faraday Cages....



Original poster: "Jim Lux by way of Terry Fritz <twftesla-at-uswest-dot-net>" <jimlux-at-earthlink-dot-net>

> >
> >
> > I think that radiated fields (which is what screen rooms are for) are
the
> > least of your problems.  Worry about magnetic fields, and transients
(also
> > magnetic fields).  I once spent several months struggling with isolating
> > 60Hz (and harmonics) interference from EEG recordings with all manner of
> > shielding, iso transformers, all sorts of grounding. What finally worked
the
> > best was NO shielding, and careful attention to wiring layout (lots of
> > twisted pairs and NO ground loops).  You want to minimize the area of
any
> > "victim" loops.
>
> this is something that has baffled me, i have been in mcr's that the
> magnetic fields were almost palatable when 8 or 10 1200 hp motors were
> under full load, but yet i could plug my laptop into a drive and check
> faults or make timing changes? i never understood why these fields
> didn't act on the pc or even the floppy's that were stored in them? is
> the magnetic in a strong pulse more damaging then a steady, but
> fluctuating field?

The key thing for a magnetic field and interference is the rapidity of
change (the di/dt in the sending loop).  Essentially what you have is a
transformer with the current coupling from the source to the victim.  Just
as with a TC at several hundred kHz, as compared to 60Hz.  With a big 60 Hz
field, it just isn't changing all that fast, and the amount of current
induced in the "secondary" (i.e. the victim) isn't that great.  This is what
makes switching noise in power supplies such a pain.  You've got those
switches turning on and off very quickly (to improve efficiency), and that
means high di/dt, which means high signals in a receiver loop.

It won't come close to changing a floppy, because magnetic media requires
pretty high fields to change state (the property called coercivity tells you
just how strong it has to be).  And, that effect isn't rate related, but
more the peak magnitude of the field.

The interference effect is more a rate of change thing.  A small field
changing very quickly is much worse than a big field changing slowly.

Computers are fairly immune to magnetic effects primarily because the
receiving loop (i.e. the secondary winding) is pretty small area (mostly
because you design them that way so they don't radiate very well, and you
can pass the FCC tests), and doesn't "intercept" much of the changing field.

The problem with lightning (and presumably TC sparks) is that you've got
fast fields that have very high di/dt (10 kA in 1 microsecond is a typical
number for lightning.. di/dt=1E10 A/second).

I'm working up some typical numbers for a moderate sized TC and I'll publish
them when I finish.. I'm assuming a 2 m high coil, making a 2 m spark to a
grounded rod 2m high (i.e. the loop is a 2x2 m square). I need to work out a
typical topload C, and then, I can calculate the di/dt.

Then, you can figure out how much current would be induced in a loop at a
given distance.