[RFI] Headphone Hum
jim at audiosystemsgroup.com
Tue Dec 10 12:40:00 EST 2013
On 12/10/2013 2:36 AM, Gary Smith wrote:
> The problem is a hum generated when I interconnect the sound card & K3,
Do you hear HUM (pure 60 Hz), or do you hear BUZZ (triplen harmonics of
60 Hz)? The causes are VERY different, and so are the solutions.
Pure 60 Hz interference is the result of MAGNETIC coupling of stray
magnetic fields into the interconnecting shield wiring. Two common
sources of these fields are 1) leakage flux from big power transformers,
like those in a ham power amp; 2) wiring errors in your home power system.
BUZZ is the result of leakage current flowing on the "green wires" of
power wiring (the third pin). It consists of harmonics of order divided
by 3 -- 180 Hz, 360 Hz, 540 Hz, 720 Hz, and those harmonics are
generated in 3-phase power distribution systems. Even though most of us
don't have 3-phase power to our homes, nearly all power distribution
outside (on the street, poles, underground) IS 3-phase.
The IR drop on the green wire creates noise voltage on the chassis of
equipment, and because every green wire has a different length, and
every piece of equipment sees different leakage current, there is a
DIFFERENCE between every chassis, and there is also a difference between
every chassis and another ground connection (like the rods outside your
The cure for BUZZ is quite simple, and nearly free. It requires two steps.
1) Get power for all interconnected equipment from the same outlet, or
from outlets that share the same green wire (in the same multi-outlet
box). If you must use more than one outlet box, bond those outlet boxes
together with short, fat copper. A length of copper braid stripped from
a junk piece of RG8 or RG213 is roughly equal to #10 copper, and works
2) Bond from chassis to chassis of every piece of interconnected
equipment with short fat copper. The chassis of laptop computers can
usually be found at the retaining screws of D-connectors.
Step #1 reduces the difference in the IR drop in the green wires to the
length of the power cord between the equipment and the outlets, and is
typically good for about 20 dB of buzz reduction. Step #2 shorts out the
smaller difference that remains by simple Ohm's Law, and is typically
good for another 20-30 dB.
Not long after I moved to CA in 2006, I was invited to guest operate at
N6RO, which at the time had buzz problems that were truly awful. The AC
distro for Ken's shack had been done VERY well, and I found nothing
wrong with it. The problem was that power for gear for the six operating
positions were taken at random from lots of outlet strips, which in turn
were randomly plugged into different wall outlets. I solved the
problems in an afternoon by doing nothing more than what is outlined above.
Another suggestion. The voltage difference between interconnected
equipment is a problem for two reasons. First, because we are using
UN-balanced wiring, and second because lots of equipment is built Pin
One Problems. Pin One Problems are excited by SHIELD current, so if
there's no cable shield, there's no shield current.
Cable shields do NOT provide magnetic shielding at audio frequencies.
Unshielded twisted pair cable connecting balanced output circuits with
balanced input circuits is VERY resistant to hum, buzz, and RFI.
CAT5/6/7 in all its variations is excellent unshielded twisted pair.
The Mackie HR824s are excellent loudspeakers (I've used a pair in my
small professional mixdown studio for many years), and they have
balanced inputs. I don't know whether the EMU has balanced outputs, but
it it does, I suggest that you use them to drive the Mackies.
Like most powered loudspeakers, the HR824s are quite susceptible to RFI
at certain frequencies. A 5W 2M talkie at 10 ft can cause full
deflection of the woofer. They are much less susceptible at HF, so I
suspect the root cause is a layout issue on the circuit boards. .
Note that the simple bonding protocol I've outlined requires more and
shorter bonding between equipment. This flies in the face of tired, old
thinking like "ground loops" as the cause of all ills. Nothing could be
farther from the truth -- the whole idea of a "ground loop" is false,
and based upon false logic. Loops are NOT the cause of hum and buzz - AC
leakage currents and unbalanced wiring are the root cause.
The ONLY interference coupling mechanism involving a LOOP is MAGNETIC
coupling, so we're back to that leakage flux.
Magnetic coupling is directly proportional to the loop area of the
wiring involved, both on the sending and receiving end. Thus, magnetic
coupling can be reduced by compressing the loop (running bonding
conductors in close proximity). It can also be reduced by rotating
either the source (the power transformer) or the receiving loop so that
the field is at right angles to the loop, and by moving the source
farther from the victim.
There are several tutorials about all of this, including several about
"The Pin One Problem," on my website. k9yc.com/publish.htm, and on Bill
Whitlock's Jensen Transformers website. Bill and I are Chair and
Vice-Chair of the AES Standards Committee's Working Group on EMC, and
principal authors of all AES Standards on EMC. Bill's work on analyzing
a balanced interface as a balanced Wheatstone bridge caused the IEC to
re-write their Standard on the definition and measurement of CMRR. It
was Bill who opened my eyes to leakage current as the cause of buzz, and
I've applied that to our ham shacks in the Power Point pdf about Ham
Interfacing (on my website)
73, Jim K9YC
More information about the RFI