John,
I'm converting from a tube amp to a solid state amp. I've seen
references to increased distortion caused by a solid state amp when the
SWR is high, not high enough to cause reduced (foldback) power. Can
anyone point me to hard numbers (not folklore) on this?
I don't think anybody can put a hard number to the change in distortion
caused by a a specific SWR. Too many factors are involved. Starting
with the fact that a specific SWR can be caused by an infinite number of
different load impedances.
Probably the strongest increase in distortion will happen when the
amplifier is underloaded. For example, imagine the amplifier being set
up injto a 50 ohm dummy load, with the drive being adjusted so that it
operates at nominal output power and distortion, and then this amplifier
is operated into an antenna that happens to have such an impedance that
the amplifier proper (before the low pass filter) is loaded with 60 ohm
instead of 50.
Since the drive remains the same, and transistors are controlled current
sources, the transistors will try to push the same current into that 60
ohm load. This would result in a 20% higher output voltage, which would
make the amplifier clip the signal, and thus generate heavy distortion.
Almost all solid state amplifiers employ negative feedback, which tends
to counteract this voltage increase, but never completely. So in
practice some light clipping will result, with some light increase in
distortion.
Note that this is exactly the same with tube amplifiers. If you
correctly load a tube amplifier into a dummy load and then connect it to
that non-1:1 antenna, the tube will clip the signal at least as badly as
the transistors. The advantage of tube amps in this regard is simply
that ALL of them have the antenna tuner built in, in the form of a Pi
tank. You need to load a tube amp into the actual antenna, not a dummy
load, and on the free frequency nearest to your intended operating
frequency, to get correct performance. If you do the same with the
antenna tuner between a solid state amp and the antenna, you are on
equal grounds and don't need to worry about this issue.
There is one difference between tubes and transistors, in terms of
distortion: It's the fact that transistor capacitances change strongly
with instant voltage. This can add additional effects and distortion
with transistors operating under non-perfect SWR conditions. But modern
solid state HF amplifiers tend to use transistors rated up to VHF and
sometimes UHF, and their capacitances are low enough to reduce the
practical effect of the capacitance modulation. Basic linearity problems
such as saturation and inaccurate biasing cause far more trouble, and
these are the same for tubes and transistors.
So, an equivalent to a tube amp with Pi output is a solid state amp with
built-in antenna tuner. A solid state amp without antenna tuner has no
real, practical equivalent in the tube world, due to the impossibility
of making broadband impedance matching circuits for the high and very
reactive impedances of tubes.
If anyone wants to run a normal solid state linear amplifier into
practical antennas without using some sort of antenna tuning, the
amplifier should use a drive control system that doesn't act on forward
power, but separately on transistor voltage and current. The other (and
better) option is to use a nonlinear (saturated) amplifier with supply
voltage modulation and current monitoring. This allows maintaining very
low distortion and high efficiency into variable loads.
If you use a commercially made solid state linear amp without antenna
tuner, I would suggest to watch the SWR and reduce the drive power as
the SWR rises. And even better than that, if you have a monitor scope
watch the waveform for clipping and adjust the drive accordingly, or if
you have some sort of spectrum analyzer (such as SDRs offer) watch your
IMD sidebands and adjust drive to keep them low enough.
Manfred
========================
Visit my hobby homepage!
http://ludens.cl
========================
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|