Robert said:
now for the 'non-scientific' answer.
ANY device that constantly reminds the wearer will have a positive
effect.
so magent, copper, woven seaweed, doesn't matter what it's made of.
Just the fact that the wearer is constantly reminded will produce
positive 'anecdotal' information.
We have inductive loops attached to the walls, spaced closely to form
an array. These loops are all driving via Atmega processors to generate
a pulse and fast compactors attached to the inputs.
They daisy chain together and the end unit will connect via serial to
the PC.
The program in each does not generate its own ID however, the values
for each zone are appended to a packet and the size of the packet and
CRC are recalculated. We have an input assigned to the uC indicating
it is the first unit via jumper, so it'll start TXing immediately and
the others will wait for a packet to arrive and then send it to the
next, etc.
The first unit will pause sending for the maximum time it would take
all 16 devices to send if they were implemented or if some error
resulted in the link some how.
With this kind of configuration, you are able to see moving objects as
they pass by the wall.
We don't use a BFO method because it is far to unstable but found that
using time constant on L works very well. The mosfet switches are used
to energize the coils. THe gate is driven via a bjt from the uC for the
on state and for the off state a pull down R is present at the gate to
allow for a slow turn off, we did this to avoid the wheeling voltages
and not use diodes, so far it seems to work
I use the atmega due to high IO count, so each uC can accumulate many
zones which will keep the lag down.
Also to add to that, when sampling in the group of IO's we need to
stagger loops, meaning, we can't energize a loop adjacent for obvious
reasons.
There are other little details we do with this system but I won't get
into that.
Jamie