[Amps] FL-2100B input swr problem

[Jim Brown]

Here's a fundamental point that use of the word "ground" as if it's a 
single point and ideal is totally wrong and causes confused, muddled 
thinking.  EMC guru Henry Ott (Bell Labs, consultant to virtually every 
major corporation, and author of the EMC "bible" now in its third 
printing) talks about "the hidden schematic hiding behind the 'ground' 
symbol.

He also talks about keeping track of where (exactly) the current flows, 
and he's talking about ALL the current in a given path. For example, 
wiring for both AC and DC power also often carries RF noise currents. 
It's absolutely necessary that every one of those paths for AC and DC 
power be treated as a transmission line if we are to control the RF 
currents in that path.

Current flows in a loop, and we must keep track of all the sources and 
all of the paths.  Thinking about those caps -- we intend for them to 
carry current, so we must know both the source and the sink of that 
current. If we want circuit stability, or if we want to short out noise 
generated by a source, we must connect the capacitor to the point that 
is nearest to the source of that noise current.

A mistake I often see in circuitry like output networks, filters, and 
antenna tuners, is reliance on a chassis to carry return current. That 
GREATLY increases the length of the path, and thus the inductance. That 
changes the circuit's resonances, AND it generates a magnetic field that 
is proportional in strength to the area of the loop, and couples into 
other circuits in proportion to the area of their loop.  Has anyone 
noticed that SMART amplifier designers use coax (or twisted pair) for 
wiring in the path for T/R switching?  This minimizes the loop area and 
lowers the Z of the path -- it's only the Zo of the transmission line, 
in parallel with the high-Z path of the highly inductive path via the 
chassis, so no return current flows on the chassis and it all flows in 
the transmission line.

> My advice to the next person tackling this repair job is to tie the replacement mica cap ground leads to the exact same points as the OEM caps, even if you think you can arrange them in a neater and easier fashion.

Exactly right -- if, hopefully, the original designer thought about all 
of the above. Or, if he isn't that smart, he staggered around with 
different layouts and stumbled into a layout that is stable. :)  Either 
way, it's a can of worms we don't want to open unless we're prepared to 
"own" the resulting instability.

73, Jim K9YC





On 6/30/2017 8:33 PM, K2CB wrote:
> What I found was that how and where the 10 and 15m caps are grounded makes a big difference in the overall tuned circuit.