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

Re: ElectroMagnetic Compatibility and Tesla coils (fwd)





---------- Forwarded message ----------
Date: Sun, 26 Apr 1998 14:37:02 -0700
From: Jim Lux <jimlux-at-earthlink-dot-net>
To: Tesla List <tesla-at-pupman-dot-com>
Subject: Re: ElectroMagnetic Compatibility and Tesla coils (fwd)


> 
> 
> ---------- Forwarded message ----------
> Date: Sat, 25 Apr 1998 14:37:31 GMT
> From: Mike Harrison <wwl-at-netcomuk.co.uk>
> To: Tesla List <tesla-at-pupman-dot-com>
> Subject: Re: ElectroMagnetic Compatibility and Tesla coils (fwd)
> 
> Remember there's a big difference depending on whether the 'target'
> equipment is switched on or off - if off, you need to zap it with
> enough energy to cause physical damage to the semiconductors.
> However, if it's switched on, you only need enough energy to latch up
> the CMOS structures within the ICs, and the unit's power supply will
> do the rest,  if it can supply enough power. The very high di/dt of a
> spark discharge is very effective at causing latch-up, and even a
> relatively small discharge could be enough. 

Yes, latchup is a problem, but, wouldn't most commercially available
equipment make sure that a lead which could get hit by a ESD pulse would be
protected against latchup. The unprotected leads should be inside the box,
and therefore shielded. Of course, violating the shield (by using an
unshielded interconnection cable, probably) would help get the transient
into the box.

For the microcontroller system I am designing right now, this is a
significant concern, since we are modifying an existing "protected" system,
and don't want to nullify the protection. On the original system, all leads
heading to the outside world (i.e. where you can actually touch them) from
the internal 68HC11 processor are diode clamped with a series resistor, and
so, can't cause a latchup. My contemplation of this issue, in combination
with some recent playing around, started me thinking about the whole EMI
problem and Tesla coils.

So, assuming the box is intact, how does the transient get inside to couple
to the vulnerable pins? A tesla coil is at such a low frequency, you could
have a huge hole in the box and it would still be electrostatically
shielded (unless an arc goes through the hole, of course). Perhaps a
magnetic field impulse from a spark? Most of the EMI protection is for
electrostatic fields (i.e. conductive paint or fillers in the plastic,
etc.) whereas you need something with a good high permeability for good
magnetic shielding.

> Moral - always switch everything you can off before firing the TC, and
> if anything you've left on starts misbehaving, switch it off
> immediately - you may be lucky & prevent thermal damage from latch-up.

Of course, but my point is that "well designed" equipment (which most is,
these days, as far as ESD) should be relatively immune to damage.
Disturbing the operation is another story... You radiate a kilowatt of RF
and anything that is a receiver, intended or otherwise, is going to get
interfered with. That is, of course, the Part 15 issue...