> One thread that I have taken from this bilious
> discussion is
> that some of you do not realize that most of the
> commercially built Amateur
> triode linear amplifiers throughout the 70's, 80's, and
> 90's have floating
> grids, of one sort, or another. It is time to pull the
> heads out of the
> sand and realize this.
Being involved in the design process for Heath and others I
can tell you exactly why Heath and others floated the grids.
They did so because Bill Orr placed huge amounts of pressure
on manufacturers to include what he called "negative
feedback". If you look at his Handbooks, you'll find a
circuit called a "super cathode driven amplifier".
The entire concept was seriously flawed, but engineers
eventually caved in to his constant pressure. Bill put as
much pressure on me as he could to continue using floating
grids in the Heathkit and Ameritron lines. When I
demonstrated the small capacitors from grid to ground
increased distortion, made amplifiers less stable, and made
them more susceptible to exciter, bias, and input network
damage....plans were made to drop the floating grids at
Heath. An amplifier tentatively called the "Warrior II" was
proposed and Heath had the prototypes I developed, but at
that time Heath was not sure if it was going to stay in the
Amateur market.
The idea for floating grids actually came from Collins.
Collins had a VERY good idea in the 30S1. Since the tube was
a tetrode with a zero control grid current rating, the
system worked very well. The screen was directly grounded
providing an arc barrier and RF shield. The control grid was
floated through small capacitors. Since the control grid
NEVER had grid current the cathode/grid voltage division
was uniform. The instant the grid drew current the grid
cathode impedance dropped like a rock. This immediately
clamped the grid to the cathode, and the tube was instantly
protected for excessive control grid current. This system
worked fine in a tetrode. It was a piece of engineering
artwork!
The problem came in when Collins applied that system, one
that worked in an AB1 tetrode, to AB2 triodes. In an AB2
amplifier the grid/cathode resistance goes all over the
place. The small mica caps no longer form a capacitive
divider that is linear with frequency and drive, and there
is no screen to serve as RF isolation or as an arc barrier
to protect the exciter and other cathode connected systems.
This unreliable feedback is why the 30L1 went through such a
long series of stability mods, and why Collins wanted a long
coaxial cable back to the exciter.
Unfortunately Orr copied that idea, renamed it, never really
grasped the problems, and successfully managed to get
several manufacturers to copy a system very easily proven to
be bad news.
An example of stability issues was the SB-220 series. I
could apply certain load impedances at LF that would make
the amp oscillate below the AM BC band. This was because the
grid RF chokes would resonate with the parallel capacitors,
and since the output network was a lowpass (pi) a suitable
length of transmission line or LF load impedance would tune
the anode to the same frequency! The only stability problem
in the 220 I ever found was below the BC band, and it was
caused by the floating grids parallel L/C combination.
This is the stability problem Rich tries to turn into a VHF
problem by saying I saw SB-220's oscillate in lab tests. It
was NOT VHF, it was LF.
As for grid protection, Heath ate a lot of tubes in the
SB230. The internal argument was that users only need watch
grid current to have good tube life, and that a positive
electronic grid overload shutoff was an unnecessary expense.
When that proved incorrect plans were made to rework the
SB-230 (including 160 meters) and add a protection system.
There were actually several amps in the works at Heath when
they pulled out of the market, including a solid state high
power amp.
All of this is exactly what happen inside the manufacturing
circle. The entire floating grid mess was created by a
flawed design that was copied and forced down the throats of
manufacturers. It only stopped at Heath after lab
demonstrations proved it to be a very bad idea, but by then
most companies using it (including Heath) were leaving the
market.
People are free to do what they want and think what they
want, but this is what actually took place inside the
industry. There are many hundreds of testing and evaluation
hours behind all of this, including statistical analysis of
failures in the field. The data base was very large, and the
goal was to reduce field problems and improve existing
designs.
I can say with absolute certainty if a grid is floated the
odds of exciter damage increase several times over a
grounded grid, and the resistance does nothing at all to
protect the most common types of tube damage.
73 Tom
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|