First, RF bias switching does not cause problems when properly
implemented. So that is a non-issue.
Second, while Rich focuses on the electronic bias with empirical
conclusions many rigs have horrid leading edge distortion.
I owned a 775 DSP for a day, and was amazed to see it produced
over 350 watts peak power on leading edges even when the power
control was backed down to 70 watts. My IC-706 does the same
thing, but at a reduced power of only 150 watts on the overshoot.
The key to successful auto-bias is NOT pulling the tubes up into
cutoff, but letting them rest at low dissipation on the class AB side
of B.
The second key is a bias switch that actuates at milliwatt power
levels, and that has a response time many times faster than the
rate of the fastest envelope rise time permitted.
That rate is controlled by the SSB filter in the rig, not by "human
speech" going into the microphone. The SSB filter bandwidth sets
the rise time.
Just like ferrite cores, this isn't a big engineering problem if you
really understand how things work.
From: yo3ctk@qsl.net
Date sent: Wed, 02 Jun 1999 15:03:13 +0300
To: amps@contesting.com
Subject: RE: [AMPS] Bias for SB-220
>
> Being on this thread, why one can possibly need RF-actuated electronic
> bias switching ? Could it be in order to save some power dissipated by the
> final tubes, corresponding to the amount of idle current in between the
> words? Come on! If the final amplifier needs this tiny margin, then maybe
> its cooling system is not so reliable after all.
The real problem is wasted power and heat. Letting the bias run in
a full-tilt on condition is like putting a 600-1000 watt space heater
in the room.
Not only that, many Hams whine louder than the blowers they
complain about. Amplifier sales drop like crazy when a blower
makes noise, so the end-effect is that noise takes priority over
excess cooling.
Because of that, and the fact that no matter how much air flows the
heat all eventually makes it in the operating room, anything that
can be done to reduce needless heat makes the operator (and
equipment, for the most part) more comfortable.
> Would you trust, for instance, a RF-actuated TX/RX switching scheme, as it
> is used in some VHF "brick" amplifiers ? I don't think so. Then why trust
> bias to the same arrangement ? In my view, the only sensible way is to
> interlock bias switching with TX/RX switching. Better yet, use a sequencer
> similar to what's on G3SEK tetrode boards.
Since it's possible to have a switching response faster than 50 uS,
and envelope rise time on speech is almost 300 uS, an electronic
bias system can turn on much faster than the envelope rises.
It is also unnecessary, and actually foolish, to force the tubes into
cut-off. The tube(s) can remain slightly above class B into class
AB, even when resting, with anode dissipation less than the
filament dissipation.
The end result is for a few dollars in parts you can reduce heat and
wasted power, and not hurt IMD even the slightest amount. An
additional advantage is white noise during keyup is reduced when
using separate antennas for QSK.
I use RF bias switching in all my PA's.
73, Tom W8JI
w8ji@contesting.com
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|