Grounded Chassis (long winded)

[Grant Richter]

Grounding is a very difficult subject. Around industrial and
power plant people, it's more of a religion than a science.
Here's a few simplified explanations I've heard up over the years.
Somebody correct me if I am wrong about any of this.

Chassis grounding is a required safety feature and
an extension of the grounded electrical system.

A little background - in 117 VAC US mains power,
the "neutral" line is a current return line, it is connected
to earth ground at the incoming service line. The balance of
the any conductive conduit is required to be connected to
earth ground also. The idea is if your "hot" comes loose and
contacts your enclosure, the circuit breaker opens.

If you do not connect your circuit ground to earth ground
at all, it is possible for static charges to build up and
"float" the entire circuit up to a potential that exceeds
the breakdown voltage of the power supply transformer.
Conventional wisdom (from the old guys in the engineering
department) is that you tie circuit ground to chassis (earth)
ground through a 100 ohm half watt resistor. This acts as
a static drain and limits fault currents in the event that your
earth grounds are at different potentials (which will happen if you
get a lightning strike close to the building).

If you consider electro-magnetic radiation, it has both an
electrostatic component and a magnetic component.
The electrostatic component is the one primarily addressed
by "shielding" since aluminum and other common
electronics material have little magnetic shielding ability.

In the case of an RF source inside of the enclosure,
the entire circuit forms one plate of a capacitor and the
enclosure forms another. With the enclosure at the same
potential as circuit ground, the RF could be thought to "short"
to circuit ground.

In the case of an RF source outside of the enclosure,
the inductance of the ground line prevents the RF from
being completely eliminated (relative to earth ground).
However, with the chassis enclosure connected to circuit ground,
the entire enclosure and circuit fluctuate at the same level relative
to earth ground. Since the ground reference is fluctuating at the same
level
as the rest of the circuit, the RF appears "common mode"
and so is "invisible" to the circuitry inside.

So (as I understand it), keeping the enclosure at the same
electrical potential as the ground reference of the circuit inside
is more important for RF suppression that the enclosure is
actually connected to an earth ground.

----------
> From: Martin Czech <martin.czech at intermetall.de>
> To: synth-diy at mailhost.bpa.nl
> Subject: Grounded Chassis
> Date: Tuesday, June 29, 1999 6:02 AM
> 
> If some circuit is housed in a unearthed metall chassis, what does this
> mean for interference?
> 
> I think the rf H-field will be weakened by eddy currents.  This may be
> also true for the E-field component.
> 
> The coupling of E and H field should be lower for lower frequencys, in
> this case the antenna/chassis is one capacitor, the chassis/ground plane
> (earth) is another, i.e. a series capacitor circuit.
> 
> Will the E-field be weaker inside the chassis?
> 
> If the chassis is grounded there should be no E-field inside.
> OTOH grounding means power cords, this will give some inductance.
> You can't really hard ground anything for higher frequencys.
> 
> The power cord may also have common mode dirt, how can we get rid of
> that without ground?
> 
> And: can we hope to get rid of that even with ground (and ground line
> inductance)?
> 
> However, it seems to me that a ungrounded metal chassis is pretty useless
in terms
> of interference. But that's what the industry seems to sell most.
> 
> 
> Any rf people here ?
> 
> m.c.