[Fwd: LF: Antenna Matching]

Andre Kesteloot akestelo@bellatlantic.net
Thu, 05 Mar 1998 17:53:02 -0500

drassew2@interalpha.co.uk wrote:

> Further to my comments last time on false match positions and those from
> G3PLX, I have made a few further changes and have now achieved a much
> better, if somewhat surprising, result.   After the correct tuning point had
> been found as detailed in my previous note, the amplifier could still only
> deliver around 80 Watts before instability set in - showing up by the
> ferrite matching core singing loudly.  The mechanism here, I suspect, is
> that the harmonics generated in the amplifier at saturation hit the
> transformer which is reactively loaded at these harmonic frequencies.  These
> components get reflected back into the amplifier and this, being feedback
> stabilsed for audio frequencies, is unstable at the 100s of kHz being
> generated.  The whole system then tends to oscillate at a frequency
> determined primarily by the ferrite core and amplifier output components
> which here is around 2 - 3 kHz.  Fortunately this oscillation is not self
> sustaining, being only maintained whilst the amplifier is being overdriven
> at 73 / 137kHz.
> Since the main instability was at a few kHz, I reasoned a high pass filter
> could reduce this effect and tried a 0.22uF capacitor in series with the
> amplifer output, in the primary of the matching transformer.  Result - at
> 137kHz - 150 to 170 Watts with the amplifier running cooler than ever before
> at this power level!  Presumably, the reactance of this capacitor (5.3 ohms)
> is being tuned out by either leakage reactance of the transformer or
> reflected off - tuning of the antenna.  What is even more surprising is that
> this capacitor gives improved performance at 73kHz as well, where its
> reactance is now 10 ohms.   It could be that the amplifier itself has a
> reactive output impedance caused by imperfect feedback at these frequencies.
> Maplin Amps vindicated!
> Andy  G4JNT