[Elecraft] K3S Audio buzz on transmit, especially AM

[Jim Brown]

A pro audio colleague, Neil Muncy, now deceased, specialized in chasing 
down hum and buzz in home studios. One mechanism he ran into several 
times was "triplen" harmonic current on the incoming neutral or ground, 
and I worked with him on one of them. These harmonic currents are the 
result of distortion of the current waveform on power lines,which in 
turn is caused by current in anything with a capacitor input power 
supply drawing most of the current  at positive and negative peaks of 
the AC cycle. Most power lines outside our home are some form of 
3-phase, and in these systems, any harmonic divisible by 3 ADDS in the 
neutral rather than cancelling. The most prominent components are 180 
Hz, 360 Hz, 540 Hz, 720 Hz, and so on, and they are heard as "buzz" 
rather than hum.

Few of us have 3-phase power in our homes, but a power distribution 
system called "high leg Delta" or "wild leg" is VERY widely used in 
neighborhoods where the customers are a mix residential and light 
industrial.  This system is a Delta, but with one leg of the Delta 
center-tapped. The 3-phase customers are fed from the Delta, and 
single-phase customers from the center-tapped transformer. Single phase 
customers get phase-phase-neutral. The 3-phase customers get no neutral, 
but they generate lots of triplen harmonics that go to ground via the 
neutral bond in single-phase homes.  My home in the Santa Cruz 
Mountains, is fed single-phase power (120-0-120) by high-leg Delta.

Depending on lots of factors, that triplen current can get pretty large, 
and the magnetic field can couple into any magnetic loop, into any 
unshielded transformer, and into magnetic guitar pickups! That's where 
Neil came in.  How strong the pickup is depends on the routing of the 
neutral and ground wire carrying that triplen current. Another colleague 
working on the design of audio for a major TV studio in NYC found very 
high triplen currents in a vertical riser in the high rise building 
housing the studio, right next to the main mix console. MAJOR problem -- 
it got into almost all the gear and low level mic wiring in the room!

As to the accuracy of your FFT frequencies -- FFT's are computed in 
"bins" with their width dependent on the FFT parameters for any given 
measurement/display width. If bins are fairly wide, the frequencies 
you're reading could easily be triplen harmonics.

73, Jim K9YC

On 4/26/2019 3:38 PM, Nicklas Johnson wrote:
> they're more or less 120Hz apart, but they start at
> a weird place-- 350 isn't a multiple of 60 or 120, but that particular spur
> and the 470 and 830 Hz spurs are all pretty exact, though of course it's
> not impossible that there's some FFT error in determining those frequencies
> and they're actually 360, 480, and 840.