Franc Zabkar wrote:

Eeyore put finger to keyboard and composed:
Franc Zabkar wrote:

This month's Australian Silicon Chip magazine has an article on the
hum and RF noise produced by 2-pronged DVD players. The author(s)
tested quite a few of these players by connecting them to an old
3-pronged amp. They found that audible hum was introduced when earth
current flowed from a player's switchmode PSU to the amplifier's earth
via the analogue signal leads.

100% predictable. The usual cuplrit is the Y caps in the SMPS of the 2 wire
equipment which results is significant (but legal) leakage current of up to
around 2 mA. Of course, this won't happen if the amplifier is a 2-wire type too
since there's no earth for the current to flow to.

The same article measured the chassis potentials of the various
devices. One measured 81VAC with respect to earth, others measured
115-117VAC. In some devices the A-E and N-E RFI suppression caps were
not installed. However, all appeared to have a coupling cap between
the AC and DC sides of the switching transformer.

That one's a killer. The idea of it is to return switching noise on the output to
the low impedance of the bulk filter cap. Of course it also results in the high
frequency content of rectified AC being passed to the secondary side.


One end of this cap appears to be connected to the case, while the other end
probably goes
to the negative side of the bridge.

Or positive. Doesn't matter which especially but returning it to the negative side
can interfere with the regulation circuitry operation.


I would think that it would still be possible for a noise current to
flow from one case to another if the chassis potentials were
significantly different.

I suppose that must be possible but maybe it'll be of a smaller magnitude.

Graham