Dear Mal, LF Group,
G3KEV wrote:
Tnx Jim for reply, but since a class D with a resonated output L/C becomes
a
single frequency tx it is getting close to class E design.
All of these circuits which depend on a tuned tank circuit of some sort
(class C, D, E, F...) are essentially single frequency devices, although
often the tank loaded Q can be fairly low, and so accomodate a modest
frequency change without re-tuning. Differences between class D and E are:
-Class D designs of the "Decca" voltage-mode type are "zero-current
switching", i.e. the output devices switch the voltage when the load current
sine wave reaches its zero points, which reduces switching losses. (the
push-pull G3YXM/G0MRF circuits are more-or-less the complementary,
current-mode, zero-voltage switching type). Class E cleverly arranges things
so that both voltage and current are zero when switching occurs, further
reducing switching losses.
-Class D are symmetrical circuit topologies with symmetrical waveforms at
the output, while the basic class E circuit is single-ended with
asymmetrical waveforms.
It would be interesting to observe the switching wave form while adjusting
L/C for resonance instead of just peaking for max output.
As Andy says the Decca-type switching waveform is ideally always a square
wave, irrespective of load impedance. The load current, observed with a
current transformer, is always virtually a sine wave. The effect of
mis-tuning is that the voltage and current waveforms get out of phase with
each other.
Cheers, Jim Moritz
73 de M0BMU
|
