Hello Nicholas,

The behaviour you describe would not pass the UL test that every alarm panel
(controller) should in the US. The tests not only include RFI, but also
direct stimulation by 60KV in both the power and telephone connections. If
the controller mistakes an alarm, misses an alarm or dials erroneously, it
may not be marketed with the UL seal. Obviously a lock up after the attack
would also fail the unit.

We have done two major premises alarm systems and they are among the highest
pain-in-the-butt factor projects we have done because of the strict testing.
In order to test the first one we did, we drew a 12" arc from a tesla coil
directly onto the functioning PIC. Not very scientific, but some test!

Tom

>
> I have tried it.  It not only trips every zone, but gets
> enough noise into
> the shielded, grounded controller cabinet that the controller
> goes into
> orbit.  After a dose of the high frequency pilot arc, the
> alarm system LCD
> displays will produce heiroglyphics and the control unit will
> randomly dial
> the phone line until it is rebooted.  Needless to say, after
> a quick exposure
> to the interference, the system is no longer capable of performing its
> intended
> function.
>
> Grounding of the welder and workpiece reduces HF radiation,
> solving the
> problem.
>