[Amps] The Pin One Problem

Paul Christensen w9ac at arrl.net
Fri Jun 7 14:32:03 EDT 2013


> 'The problem is that it is not a "Ground loop"...'

To Joe's point, a ground loop isn't even necessary to create a "Pin 1" 
problem.  The problem always manifests when current induced on a shield 
modulates a signal ground.   Few manufacturers know how to manage equipment 
and interconnect grounding.  Typically, they see one schematic ground symbol 
and conclude that it's fine to tie grounds anywhere in the equipment where 
there's continuity.  But in doing so, current splits in uncontrolled ways 
along circuit boards, ground planes, the chassis, shielded wiring, the 
third-pin of an AC line cord, etc.  In our field, the guys who really get 
the concept (notwithstanding K9YC) are folks like Rick Campbell, KK7B, whose 
high-performance direct-conversion designs demand near-perfect ground paths 
and this is discussed in his book with Hayward and Larkin.

Some of the best designed video/audio equipment have several ground layers 
including analog ground, digital ground, separate ground buses for +5V, 
+12V, +15/-15V, etc.  But every one of these ground references terminates at 
one point in the equipment, bonded to the chassis and often near the AC/DC 
line input.  In breaking up ground paths, high current ground points can be 
diverted from low-level signal ground paths.

In the typical Pin 1 instance, a cable shield is tied to circuit ground 
(either out of ignorance or cost/convenience) and not bonded to the chassis. 
If there was perfectly zero resistance between the chassis and *every* 
circuit ground point, we wouldn't have a problem.  But just small amounts of 
resistance (e.g., a PCB trace) can create enough of a voltage division to 
affect low-level stages.

No matter how many words are used to describe the problem, I think one of 
the best teaching tools come from an illustration.  In Paul Horowitz' "The 
Art of Electronics," he shows a right and wrong way to approach equipment 
grounding.   No doubt K9YC's website has at least one example and a picture 
speaks many words.

I don't like the "Pin 1" term, but there may not be a better way to describe 
the problem in a one of two word title.

Paul, W9AC







More information about the Amps mailing list