[Elecraft] KIO2 noise update

[Jim Brown]

On Wed, 24 Dec 2003 09:52:33 -0800, Scott T Emery wrote:

>
>	I then measured the case ground and saw a similar wave.  I didn't
>believe this, so I measured the difference between the scope ground
>and the computer ground with a DVM using and came up with 32VAC. Something
>is wrong *HERE*.
>
>	To explain things a little better.  There are two outlets in
>the computer room.  One has a Power conditioner feeding some PDUS
>feeding some computer related equipment connnected to it.  The other
>was conveniently unused.   I saw the difference in ground potential
>when I connected the O'scope to the unused outlet and measured the
>computer system.   Plugging the test system (and the K2) into the
>computer's PDU made the problem go away.

This is a very common problem. We encounter it often in pro audio and
video systems, and take care to install dedicated isolated ground for
them. 

There are several possible  causes of your problem. One is a miswired
AC outlet (phase and neutral reversed). Another is a mis-wired power
system. 

A common fault is the improper connection of the AC neutral at more
than one point. By code, neutral must be connected to ground at one,
and ONLY one, point. That point is usually the secondary of the
transformer that generates the circuit. But many panelboards (breaker
panels) have a bonding screw that, by default, ties the neutral bus to
the frame. If the neutral is already bonded at the transformer, as it
should be, this default bond at the panelboard is a violation of code,
and must be removed. More important, this extra bond generates
circulating currents on the ground conductor, which causes voltage
drops. 

Another contributor to the voltage difference between "grounds" at
different locations is the IR (and IZ) drop from current that flows
when bypass capacitors take noise to "ground."  MOV's do this too. The
far better way to do surge protection is a series mode suppressor like
those made by SurgeX. www.surgex.com That's what we do in the pro audio
world. (No, I don't sell them). They don't short the spike to ground,
but instead store it in a big LC network, then release it slowly back
to the line. 

Also, in a 3-phase system, triplen harmonics won't cancel on the
neutral, but will add. In a building with mostly non-resistive loads,
it is quite common for neutral current to be 2X the current on a phase!


It is quite common for the potential difference between neutral and
ground at an outlet to be a few volts. You should not see O volts (that
indicates a neutral shorted to ground), and you should not see a lot
more than a few volts. 

The solution you found is a very good one -- power all your station
gear from the same outlet. 

Jim K9YC