In a message dated 96-11-19 23:18:05 EST, you write:
>
>Is a Q of 8 enough to provide sufficient harmonic rejection/TVI
>supression, etc? All of the "how to build a linear" articles I've seen
>over the past 40 years of the magazines (I had a fair size collection
>before I moved the last time) talk about a minimum Q of 10 and preferably
>12. Even with the "high impedence" tubes, 1 + (Rplate/Rload)^.5 gives a Q
>of less than 9 at amateur power levels.
>
>Of course, keeping the Q down to 10 is a real trick at 10 meters with the
>output (plate to grid) capacitance of most tubes. The MLA is a real
>problem with it's 75pF minimum in the tuning cap!
>
>--
>
>73,
> ... Joe Subich, AD8I
Hi Joe,
The question was efficiency, but I thought about what others might think
after I posted it.
Perhaps in hindsight should have said more. I suspect the Q of 12 comes from
days when PA's were often class C, and is a ballpark figure that crept in as
an accepted absolute requirement. Of course a class C PA is pretty ratty, and
with 1500 watts out would likely never meet specs even at a Q of 12 or 14.
But it might with the old specs of ~700 watts out (1000 dc in).
***Anyway, the point I wanted to make was a lack of Q does NOT lower
efficiency...it simply (at the point of the formula I mentioned) means the
network won't act like Pi anymore.*** Efficiency is always higher with less
Q, so long as minimum Q requirements are met.
I've found most AB PA's exceed FCC specs with minimum Q, and when they do
require extra "clean up" it's usually a problem a 30 % Q increase won't
solve. Tank layout means much more than loaded Q in many cases, and we won't
talk about the layout, hi.
I ran a Chaffe analysis on that PA years ago. I just got back in town and
will try to locate some old data I had. As I recall one of the major
problems with the MLA was not enough delta C on 160, because of all the heavy
padding required. Making tank Q higher than the minimum will give very little
harmonic improvement but make the operating range of the tank even more
restricted. Too much fixed C, and not enough variable C means anything you do
will be a compromise.
73 Tom
--
FAQ on WWW: http://www.contesting.com/topband.html
Submissions: topband@contesting.com
Administrative requests: topband-REQUEST@contesting.com
Sponsored by Akorn Access, Inc & KM9P
|