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

RF safety



A couple of days ago, I attended an RF exposure safety class at work. A lot
of interesting information was presented, particularly with respect to tesla
coil operations.  I had an opportuntity to talk to the instructor (who is
quite knowledgeable about RF effects on humans, etc.) about this in some
detail.

The significant figure to remember here is that NO repeatable effects have
been noticed at less than 3-4 Watts/kg specific absorption rate.  The ANSI /
FCC / OSHA standards are based on limiting occupational exposure to 1/10 of
that, i.e. 0.4 W/kg. For the uncontrolled general public, the limit is 1/5
of that.  Coilers would fit in the occupational category, I should think.

At low frequencies (i.e. <3 MHz), the limits are formulated more to limit
the conducted or induced currents

Some general comments:

1) At TC frequencies (hundred kHz area) ohmic heating will be the primary
concern.  The varying field will induce currents in you  (just as it does in
any conductor), those currents will result in I^R heating. In other cases
where people have had significant induced current dosages (typically such as
working around aircraft on the surface of a carrier near transmitting
antennas), the primary symptom is heating, discomfort, or sensation in the
wrists and ankles.  The current path (capacitively coupled) is through the
hand, then the body, then through the legs to the ground.  The wrists and
ankles have the highest resistance (lots of bone), so the most power is
dissipated there.

2) The ANSI specs (IEEE C95.1-1991) provide a way to relate E or H field
strength to power density.  For TC frequencies, the maximum field is 614
V/meter (or 163 A/m for H field).  A program like Terry's Efield will allow
you to calculate the efield around a TC.  You can also do a quick
approximation: a 1.5 meter coil with 1 Megavolt on it (typical numbers)
would have a field strength of 660 kV/meter (roughly 1000 times the ANSI
limit).  However, the field drops off fast as you move away.

We'll make an assumption that the field is due to the topload considered as
a point source against an infinite ground plane.  We'll also make the
assumption that we're interested in the E field on the ground (actually, on
a line halfway between the point source and it's image)

Call the height of the topload above the ground "H"  (That is, the distance
between the point and its image is 2H.) Call the distance from the line
between the topload and ground (how far you are from the coil), "R".  The
voltage on the top load is V

The E field will vary as Emax/sqrt(H^2+R^2)  (note that at R=0, the field is
Emax=V/H)... So, using our example where the 1.5 meter high coil has a field
of 660 kV/meter.  To get the field down to to 600 V/m, then 1/(H^2+r^2)
<.001, or roughly, R has to be 30 H... 45 m away!!

But wait, there's more.. you can average over 6 minutes, and you can take
the duty cycle into account.  A typical disruptive TC actually only has the
peak E field for a very short time, and then has several milliseconds of
dead time.  A duty cycle of 10% or even 1% might not be a bad estimate for a
typical TC, when actually making sparks ( for a loaded Q of 10, at 100 kHz,
the RMS value of the first 100 cycles (1 mSec) is 0.22 times the peak value
of the first cycle, and an additional cycle only raises that about .001)..
So a reasonable RMS power estimate for a synchronously fired tc at 120 Hz,
with a 100 kHz fRes is about .026.  This means that your integrated field
limit will be reached at a much closer distance.  1/(H^2+R^2)<.05... or R
about 4H or 5H, that is.. 6-7 meters... a much more reasonable number.

If you are running a DC coil or running a high break rate, you'll have to
take this into account... Increasing the break rate to 1000 pps (from the
120 pps, above), would significantly increase the RMS field.

The other issue to consider (if you want to flog the deceased equine) is the
field changes due to the sparks.  And, a coil that isn't breaking out could
have significantly higher RMS fields (because the Q is much higher without
the load of a spark, so the duty cycle is higher...) This might account for
the anecdotal observation of more RFI when the coil is out of tune......
more power is actually radiated, instead of being dissipated in the spark.

Note also that if there is a conductive membrane between the observer and
the coil (i.e. a chicken wire mesh) then all these field strength
calculations go out the window....


3) The other part of the ANSI  MPE (maximum permissible exposure)  limits
induced and contact RF currents to 450f mA (*2 for both feet) (f is
frequency in MHz), for f<100 kHz, and a flat limit of 45 mA for frequencies
over 100 KHz.  I haven't done the calculations yet on induced currents, so I
don't have any feel for whether this is a significant limit.








Jim Lux
phone:818/354-2075  fax:818/393-6875
Spacecraft Telecommunications Equipment Section
Jet Propulsion Laboratory   M/S 161-213
4800 Oak Grove
Pasadena CA 91109