[Lowfer] Class C v Class E

[Alan Melia]

>>   One problem with Class E amplifiers is the high harmonic content of
the output so adequate filtering is essential if you're going to put
something on the air.

Hi all and Ed, it all depends upon whether what you are measuring is really
operating in Class E (set Nat Sokal's QST (I think) article) The minus point
is that Class E is a "single frequency" or at least narrow band mode at LF
and also it is "single power". It is possible to operate a design at lower
supply voltages but the efficiency degrades and of course the output
impedance hence the matching to any filter changes.

 To be really Class E the tuning must be accomlished to be nearly signwave
at the switch Drain (Collector or even Anode!) This means that the THD is
actually quite low and lower than many other "switching" modes. Filters to
work (see Nat's paper again) but the problem is that like all filters they
must be matched and that means discipline in tuning. You must tune the amp
for the design power NOT maximum smoke.

Jim Tonne has a programe that evaluates the components and the there is an
Excel spread sheet on my web site. This is not just theory, several (Five at
least including VO1NA's 100 watter on 137.77kHz) successful single ended
designs have been built in the UK using the "sheet" values, from a few watts
up to about 550 watts output, for both 136 and 500kHz. We may have been
lucky but to the best of my knowledge we have not lost many FETs in
operation with these single ended designs, unlike many of the push-pull
class D designs.

Components you must use High voltage "pulse rated" capacitors of the type
specified for switching PSUs. Old high voltage wound types will not carry
the current and the connections between the foils and the termination will
fuse steadily driving down the capacitance value. Tuning sound a bind and
variometers complicated, However the inductance is not that high (around
100uH) and a simple "telescoping" pair of windings can easily be fabricted
on plastic plumbing stock, giving an adequate tuning range and be driven by
an old drive cord assembly !! The messy part is that tuning need to be done
with a scope initially at least. Efficiencies or over 90% have been seen
though as was stated on the RSGB Group reccently RF ammeter accuracy is not
that great and there is a lot of room for error. I use a Tek current probe
and a scope for accuracy.

http://www.alan.melia.btinternet.co.uk/classepa.htm

Cheers de Alan G3NYK