Amps
[Top] [All Lists]

Re: [Amps] 2m Linear Amp Enquiry

To: amps@contesting.com
Subject: Re: [Amps] 2m Linear Amp Enquiry
From: Steve Thompson <g8gsq@ic24.net>
Date: Tue, 14 Sep 2004 19:07:05 +0100
List-post: <mailto:amps@contesting.com>
On Tuesday 14 September 2004 16:03, Skram, Helge (MED) wrote:
> The "easy-build" amps by YU1AW are interesting, and at first I could not
> believe the quoted efficiencies.
> Ouput power is remarably high for such simple designs.
> The efficiency is typically in the 60-65% range.
>
> Bandwidth indicates typically Qloaded=9, which is much lower than usual
> stripline designs.
> With this low value it is possible to use a fairly lossy inductor.
>
> Looking at the output circuit design, it is similar to the K2RIW for 432MHz
> when we convert the stripline to an inductor as it really is.
>
> My problem is, how is it possible to achieve Qloaded as low as 9 at 144MHz
> ? Qloaded beeing approximately Rp/Zcout, where Zcout is the impedance of
> the tube capacitance pluss tuning capacitance transformed to the tube pluss
> spread capacitances.
> For a GS-35B:   Cout=4.5pF  When we add another 4.5pF for tuning and spread

I don't know how the capacitances for the Russian tubes are measured, but most 
manufacturers use a special isolated setup so the output capacitance is only 
that within the tube itself. I've measured several external anode types when 
mounted in a chassis and in every case the actual capacitance from anode to 
ground is at least 2x the datasheet value. I'm not suggesting that the YU1AW 
designs can't/won't work, but what's going on in the matching circuit might 
differ from the lumped element analysis. In a 50MHz amp, I've seen 
differences much greater than might be expected when changing between 
slightly different tank coils of the same measured value. The higher the 
frequency, the more wierd the effects.

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

<Prev in Thread] Current Thread [Next in Thread>