Message-ID: <000701c829fb$53914aa0$47e8fc3e@g3kev>
From: "hamilton mal" <g3kev.ham@virgin.net>
To: "rsgb" <rsgb_lf_group@blacksheep.org>
Date: Sun, 18 Nov 2007 15:54:03 -0000
MIME-Version: 1.0
Subject: LF: classs E
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org


Hi all
I carried out some analysis on my fet pa today to check the waveform for
true class E. Monitoring the drain on a scope revealed that the waveform was
almost correct but needed a slight increase in capacity for C2 for an almost
perfect switch at 502.5 khz. Increasing/decreasing the voltage did not make
a big difference to the waveform, at the design voltage it was only slightly
better.
Using a DL I altered the freq and found the perfect waveform at 535.9 khz,
so I expect logically increasing the value of C2 would correct the anamoly
at 502.5 khz. I have not altered it yet because I do not have a suitable
capicator.
I did observe some ripple on the waveform that is probably caused by feed
through from the driver, it was more pronunced at a V lower than the design
voltage, but improved with a V increase.
In spite of the above observations the PA performs well and seems very
robust.
The RF pwr out  measurements using the DL or the Antenna do not vary,
indicating that the antenna is well matched to 50 ohms via the atu at the
base of the inv L.
I thought the above observations might be interesting to would be class E
constructors.
Considering mine was built from junk box bits, including the 60 year old
ceramic ex TU5 tuning unit coil  it works very well. The fet is out of an
old marine tx.
Best dx to date VE1ZZ at 549 during ideal conditions last week.
73 de Mal/G3KEV