Amps
[Top] [All Lists]

[Amps] PARALLEL CAPS IN OUTPUT and using Mr. Smith's chart

To: amps@contesting.com
Subject: [Amps] PARALLEL CAPS IN OUTPUT and using Mr. Smith's chart
From: John Lyles <jtml@losalamos.com>
Reply-to: jtml@vla.com
Date: Sat, 07 Dec 2013 21:24:23 -0700
List-post: <amps@contesting.com">mailto:amps@contesting.com>
Fellow amplifier builders,
The kind of analysis described below, using transmission line analysis, or the shorthand version, the Smith Chart, is something that we all should be familiar with, as circuit and amplifier builders. Over the years I wrote general and problem-specific transmission line calculation programs in Basic and Fortran for my amplifiers, also using pads of paper Smith Charts. They taught those at in the radio engineering curriculum in school, decades ago.

Now there is a handy and quick way to lay down component values in a shunt or series fashion and see what the Smith Chart will show. Your smart phone and one of the free Smith Chart apps is needed, and the cost is free or just a few dollars. I prefer the Smith Chart Matching Calculator by Brett Wakley; it is fantastic in that it is simple to learn and use. It shows both schematic and the chart and allows me to adjust values as it walks around the Chart to the desired Z. I can't believe how easy it makes matching, although it doesn't have automatic convergence algorithms. When everyone else in the airplane is playing games and reading their Kindles, you can find me plugging in my favorite pi or L network and finding the best way to match things.

73
John
K5PRO


Hey guys,. this is the kind of thing that can easily be modeled and derive
some factual data.

If a 50R load is connected to the amplifier output then through a strap,
wire or coax of a given length, say six inches long what happens to the 50R
is  dependent on the characteristic impedance of the 6 inch piece.

Transmission line software will show what happens. At 28MHz, 6 inches og
RG213 would be about 8 degrees long. Also, the 50R load would still be 50R at
  the end of the line.

Changing from RG213 to a wire or strap would lower the velocity factor
which will shorten the line length to about 5 degrees and the impedance would
increase to a number determined by the spacing of the wire or strap to the
underlying ground plane (chassis). A highish impedance could be on the order
of  250 ohms for a wire that is .125" in diameter. Smaller diameter would
have  higher Z and larger diameter = lower Z. Closer spacing to ground plane
= lower  Z.

Anyway, taking the 250 ohm value, the 50R would be rotated to about 61R
shunted with about +125 ohms X. To restore proper matching in this case would
require a reduction in the load C to match 61R and cancel the +125 X.

Use of a wider strap would lower the impedance and require a smaller change
  in the load C to restore the match.

Obviously a shorter length would reduce the correction needed.

If the amplifier doesn't follow this kind of prediction then there is some
nuisance reactance present and not accounted for.

At a lower frequency such as 3.8MHz, the line length becomes so short that
a very small rotation of the R and X happens and the effect is basically
negligible.

This can also be plotted on a Smith Chart for a better visualization of
what is happening.

73,
Gerald K5GW


_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>
  • [Amps] PARALLEL CAPS IN OUTPUT and using Mr. Smith's chart, John Lyles <=