[hpsdr] class E

[FRANCIS CARCIA]

Hi all,
  Usually a class E final wants a specific drain shunt c on each band. I know you can get away with maybe 2 bands but beyond that you become class D or inefficient class E. It is actually pretty easy to build a final for each band. I would be interested in any published data out there. My concern with Penelope driving a pdm modulator is the phase shift through the pdm filter. I used a uc1846  current mode PWM in my 160 meter rig and a discrete modulator in my old Viking 2 tube TX. Some of my smarter friends put a bit of bias on the gates of a class E final to reduce the drive required. I have not tried this yet because I rely on the zero bias to assure less than 50% duty cycle drive to my push pull final. This guarantees no cross over currents due to the delay differences between turn on and turn off of the final FETs. The best cheap FETs out there today are the Fairchild FQA11N90s but I hear they are being discontinued. These are power supply FETs that work up to 40
 meters. 
  I'm working on a linear amplifier design and found positive bias with drain current requires a bit of source resistance. I just did a design for a friend and I put a pair of 1 ohm 2 watt resistors in parallel on each source to keep them linear. Also added a bit of RF feedback.
  The 11N90s seem to lose a bit of efficiency at 40 meters but still look linear at with the two forms of feedback..
  I plan to start with an old 3 stage 100 watt strip out of an old Raytheon marine rig modified to go to 10 meters. 
  The IXYS fet line was solt to I think microsemi. They are much more expensive but will cover the whole hF spectrum. There are a couple different sources of the old Motorola style FETs that drop right into CCI boards. frank 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20080216/f1065afe/attachment-0002.htm>