Turntables and Hum

David, can I start by saying the guy is selling a product. He also sells interconnect cables that do not shield-out RF.

Ever since I was a lad, there have been "problems" with dimmer switches interfering with small signals.

The earliest form of radio was spark transmission - it had no position on the dial - it was so broadband anything picked it up.

Now, inside some monolithic silicon, your lamp's power is being abruptly switched. There isn't any spark; instead, it is a reactance due to inductance and capacitance. Its transmission medium is the mains supply.

When the lamp is fully on, the Triac (or a rectified Thyristor) cannot contribute any noise. However, when dimmed, the Triac gate is "fired" to switch during a mains half-cycle.

Whereas the diodes of a rectifier cause enough problems in switching somewhere close to a volt, the bi-directional controlled silicon diode (Triac) switches power to the load at some point during the much larger potential of a mains half cycle.

The voltage can be anywhere between zero and 350 volts peak. It "spikes" back down the supply cable, transmitting distorted noise at 100 cycles per second, both airborne and mainsborne.

The occasional earthing of class-II equipment through the human body is not related as simply as some marketing suggests.

If each source item were to be mains earthed, loops would exist to inject mains frequency noise into every interconnection. This is most noticeable with turntables and less noticeable with line sources. There are different degrees of signal to ground noise ratios.

Therefore, it is crucial to either have only one mains earth connection so that a loop cannot be formed or even none at all.

There is no connection to mains earth in a fully class-II system, which has to exist in an "accessory only" system. Mains earth is what it says. It is a connection between the mains supply and terra-firma - the ground we walk on.

That fully class-II system can then connect at any point with a class-I item, such as an integrated or power amplifier, without causing an earth noise loop.

What is important is the ground continuity, such that the shield - the common conductor of the interconnect - has no resistance at signal frequencies to speak of—the two class-II items so connected become one. The case/screening of one is the same potential as the case/screening of the other; therefore, no ground noise can exist.

Suppose everything in a system is class-II (some older amplifiers just had live and neutral and no earth in their supply cables). In that case, there is no mains earth, but as explained above, there can be no ground noise (including mains buzz).

But, if we introduce the human body in the form of touching any case/screening, the human body conducts it to terra-firma - or earth. Still, no noise can happen unless the loop is completed somewhere else, and the interconnect shield allows a rise in potential between the units.

That somewhere else is usually because the accessory system connects to an earthed amplifier. However, if it results in buzz, then it is because the human body can conduct to earth more effectively than the interconnect's shield can conduct between the equipment.

That last statement might sound a little odd. Surely wire can conduct more effectively than the human body? It depends on the inductance of the wire!

Suppose the connection between the cases/screenings is inductive. It acts as resistance at some frequency, and resistance allows the cases/screenings to break company.

As I said earlier, buzz (interference) is also airborne. At the frequency where the wire becomes resistive, it becomes an antenna - a receiving aerial!

Therefore, the cases/screenings have a signal potential between them. This forms a differential signal with the music signal (or, indeed, the silence when not playing), and buzz is amplified. The transducer/s reproduce it for you to hear.

Therefore, the shield needs to be effective at all frequencies, right up to the highest radio spectrum frequencies which may be encountered. A velocity of propagation of 85%, multiplied by the frequency of light, will do nicely.

But, and this is the but, the shield must be connected at each end. It must not be subordinated to another piece of wire that completes the run - as do many "hi-fi cables." As you should be now able to see, that wire will have greater inductance and allow that differential signal to form.

Likewise, the knitted or otherwise bell-wire constructed interconnects offer zero signal protection because of the case/screening interconnection's inductance.

Understanding the above shows that if done correctly, there is no requirement for something that treats the symptoms.