AW: "gndloops"

[Haible Juergen]

	>TO SAY IT LOUD AND CLEAR: I THINK THE GND LOOP BY CABLE SHIELDS
	>CONNECTED AT BOTH ENDS IS NOT HARMFULL IN THE SPECIAL APPLICATION
I'M
	>TALKING ABOUT.  IT IS USEFULL. That's why the hyphenation in the
	>subject field. (Ok, if there is a strong magnetic ac field, it will
	>induce a current, but this should be avoided anyway).

The omnipresent magnetic field (50/60Hz) is the bad guy, IMO.
The point is that a magnetic field does induce a *voltage*, not
a current, in a closed loop.

The Maxwell equation to look at is   rot E = - dB / dt.

If you solve this along a closed loop of conductive material (such
as a cable shield, plus the internal gnd connections), you see that
there is a fixed *voltage* induced across this loop, which only
depends on the magnetic field change and the loop area.
This causes some remarkable current in the loop, according to
Ohm's law, of course. But the current isn't really that important.
Important is the impedance of certain parts of your gnd connections
around the loop. The higher a certain impedance (relative to the other
ones involved), the higher the voltage that will appear across this 
specific section of the gnd path. All voltages across the loop will 
add to the total induced loop voltage of course.

Now the bad thing about this mechanism is that if you make a very
good (low impedance) shield connection on both sides of the cable,
most of this voltage will appear anywhere else across your synth's internal
ground paths, and will result in hum there. If you simply open the shield
at one end, you'll probably get some noise voltage here, but it might
not do much harm because of the high signal levels. But with a GND loop,
most of this voltage would appear anywhere inside your sensitive circuits,
probably after the input attenuators of a sensitive ota input, or inside
of an opamp's feedback connection.

The best solution is a balanced connection, with cable shields connected
to chassis, but not to signal gnd. Then you have hum currents in your 
chassis, but not inyour circuit.

So far the theory of what I think is the most important effect of gnd loops.
The other points you mentioned are important as well.

Here's a practical hint that I have tried myself: On my Wasp Filter clone
I have several audio and gate connections (to trigger the envelope).
Sometimes I want to trigger the envelope from a different synth than
the one that provides the audio input, and so I get quite large gnd loops.
(Unfortunately most of my gear is still unbalanced.)
I tried to make use of the theory I described, and put a small (100 Ohm)
resistor into the *gnd* connection of the gate and switch trigger inputs.
(NOT the audio inputs!). The idea is that I can "fix" almost the whole
hum voltage of the entire gnd loop across this resistor. So it will 
(intentionally) affect the gate threshold a little bit (but this doesn't
really harm),
and leave the audio path alone instead.
This works great when you have a small number of connections, so you
know exactly what you're doing.

I still have to solve the problem on certain modules of the JH-3, however.

JH.