Content-Transfer-Encoding: 8bit
Message-ID: <3.0.1.16.19990819144455.3d4776e4@mail.cc.kuleuven.ac.be>
Date: Thu, 19 Aug 1999 14:44:55
To: rsgb_lf_group@blacksheep.org
From: "Rik Strobbe" <rik.strobbe@fys.kuleuven.ac.be>
Subject: LF: high power class E transmitter
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

Charmed by the simplicity (only 4 components) and the high efficiency of
class E amplifiers I decided to do some experiments with these.
As background I used a very good article by Bill Cantrell, WD5CVG, that is
available on internet at :
http://www.qsl.net/k3pgp/Notebook/Wd5cvg/Classetx/classetx.htm

A low power version running on 12V and with the antenna simulated by a
330pF (silver-mica) capacitor and a 50 Ohm resistor had 1.6W output and an
efficiency of about 96%. 
The only drawback was a rather poor harmonic surpression of about 25dB,
what would make an additional LPF nessecary (increasing the number of
components needed from 4 to 7 or even 9) at higher power (to keep the
neighbourgs happy).

As next step I calculated the required values for a 500W version, and there
comes the disadvantage of this design :
Having a loss resistance of about 130 Ohm I will need a 336V power-supply.
For less loss the required voltage decreases but still remains high (294V
for 100 Ohm, 208V for 50 Ohm). As the voltage over the FET and
shuntcapacitor can be up to 4 times the supply-voltage I would also need a
1.4kV FET what makes the cheap power-FET's useless.
If I want to use my existing 56V/750W PSU I have to limit the load to 3.6
Ohms (to get 500W power), what means that I will need a 1/36 power
transformation (either doing this in some LPF stages or using a 1/6
transformer).
But I'm afraid that including LPF (for harmonic surpression) and impedance
matching will decarese the efficiency close to values that I have right now
with my 'G3YXM-type' PA (= 85%).

So I wonder if someone already has experience with a high-power class E
amplifier ?

73, Rik


Rik Strobbe  ON7YD
rik.strobbe@fys.kuleuven.ac.be
Villadreef 14  B-3128 Baal  BELGIUM   (JO20IX)