<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.6000.16608" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT face=Arial size=2>Thanks for the most interesting discussions on PWM
for Class E amplifiers. There seems to be a number of possibilities as to how to
overcome the low voltage output problem.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>1. Use a closed loop control system that compaires
the output envelope with the modulating signal. This appears to be what Josef,
DJ7AW, does with his single transistor 80m -15m Class E 400w PA. I have a pdf of
his article, in German, that explains his complete amplifier. I can send it to
someone who could perhaps put it on a download site. </FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>It looks like the reason he stops at 15m is not
because the PA wont operate at this frequency but the fact that he can't
compensate for the phase shift between the RF drive and envelope. We have
independant control of this phase relationship with Penelope so should be able
to get the design to work on 12 and 10m. </FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>His design uses a single ARF447 which looks like a
$36.00 device. I'm not sure how well the feedback loop will perform given
the delay throught the output filter.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>2. As already suggested, use a linear regulator for
low output voltages which switches to PWM at higher output levels. This was the
path I intened to go down until seeing DJ7AW's circuit.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>3. I know of a commercial Class E PA that does it
another way. They bias the FET in Class A for low level inputs and don't
clip the drive signal until a specific level is reached. The fact that the drain
volts do not drop below 8 v is then not a problem. In effect the PA moves from
Class A to Class E with varying input signal level. </FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>It looks in this instance that the RF side of the
project is relitively straight forward and the challenging part is the analog
side.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>It would be great if we could develop a Class
E PA as an HPSDR project.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>73's Phil...VK6APH </FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV></BODY></HTML>