Topband: Chassis Bonding

[Jim Brown]

On 1/28/2014 4:17 PM, Tom W8JI wrote:
>> No transformer is necessary. Simply bond from chassis to chassis of 
>> all the interconnected equipment, and, if possible, get power for all 
>> of it from the same AC outlet.
>>
>
> That doesn't always work, Jim. 

Indeed it does. Read along below.

> As a matter of fact it is a generally bad idea, because if a 
> connection develops any resistance you wind up with hum and noise.

How does a proper 6 ft long #10 copper bond strangely "develop" resistance?

> We would have been thrown out by the ear if we were caught bonding 
> cabinets to stop unbalanced line ground loop hum or noise at a BC 
> station.

A ham station is not a broadcast station. In a BC station, equipment is 
widely separated, interconnects are BALANCED, the signal to noise ratios 
required differ by at least 40 dB. In a ham station, interconnected 
equipment is quite close together, chassis bonding can be 6 ft or less.

And in the broadcast stations where I've worked, anyone would have been 
thrown out on the ear for using any unbalanced audio connection at all!

>
> Our radios and equipment have common chassis grounds for power 
> supplies, and the currents on ground leads can be 20 amperes or more. 

20A DC, and, as we discussed several years ago, 20A DC modulated if it's 
SSB, but NOT 20A of hum or buzz.

> If the bond lead has .01 ohms resistance, there is the potential for 
> over .2 volts hum or noise.

The source of the hum and buzz is LEAKAGE current, not power supply 
current.  Leakage current rarely exceeds 10-20 mA unless some piece of 
gear has developed a fault.

>
> A transformer, on the other hand, works flawlessly. 

An unshielded transformer is a sitting duck for the leakage flux of big 
power transformers, like those in power amps. Ask K6XX what happened 
when he ran a K3 a few feet from a big power transformer using AFSK. The 
K3's input transformer (which is unshielded)  picked up enough of that 
field to regenerate!  I've heard that on a much smaller scale here with 
the power transformer for my Titans. The band-aid for that (which I 
suggested to Wayne) was to configure the DSP in AFSK mode as a steep 
high-pass with the cutoff in the 300 Hz range.

> My headphone lines, because they serve multiple desks, all have 
> isolation transformers. 

That may be appropriate if you're using centrally located headphone amps 
and/or signal distro in your station, but it's completely un-necessary 
in a single-op (even SO2R) shack.

> I suppose I could run a big copper buss bar across the room, but it 
> seems more logical and safer to just spend $2 on a transformer at each 
> radio. :)

N6RO's 6-position M/M had bodacious buzz problems when I tackled it 
about five years ago. His station was built with EXCELLENT power distro, 
with a local panelboard, with thinwall steel conduit to multiple outlet 
boxes for three pairs of adjacent operating positions. Each could be 
operated separately or SO2R. The buzz was present because they had added 
a bunch of outlet strips randomly plugged into those outlets, and 
equipment from each operating position was randomly plugged into the 
strips. When I started, I measured several hundred mV of buzz between 
the chassis of equipment at an operating position.

My solution was exactly as I described to you -- all the equipment for 
each operating position is powered from outlets that share the same 
backbox, and thus the same path to the panel for their green wire, and 
every chassis of interconnected equipment within an operating position 
is bonded together.  In several instances, interconnected gear had to be 
powered from different backboxes, so those backboxes were bonded with 
braid stripped from some junk RG8. And all the outlet strips went into 
the trash.

Ken's station may be different from yours, in that he has no centrally 
located audio distro or routing, and every power amp is next to the 
associated transceiver. It essentially resembles three SO2R stations 
each of which has equipment clustered around the respective operating 
positions. In this way, it resembles the vast majority of ham shacks, so 
the bonding architecture that works for it will work for them.

It's critical to understand the root cause of hum and buzz in our 
stations. It ls NOT "ground loops," it is LEAKAGE CURRENT from the AC 
mains supply, and the differing potentials created by these leakage 
currents in he green wires. Once you realize this simple fact, you 
realize that we can reduce those potential differences so that they are 
below the level at which they audible by the very simple bonding 
practices described.

73, Jim K9YC