[Amps] AB1 vs AB2 ???

Rich 2@vc.net
Fri, 29 Nov 2002 04:45:35 -0800


>
>Greetings Amps gurus,
>
>Happy Holidays and special thanks to all who've been so helpful to me over 
>the past year!  It means a great deal to me.
>
>I'm wondering if someone can explain why AB2 GG service is more linear 
>than AB2 grid-driven service? 

**  AB2 grid-driven has c. 20db gain - but the configuration has no 
inherient RF negative feedback to counteract the distortion products 
brought about by operating the grid in the grid-current region.   
-  AB2 cathode-driven has RF negative feedback with the tradeoff of c. 
8db less gain.  As a result, cathode-driven AB2 is roughly 15x cleaner.  
For c. 20db gain and low distortion, an AB1 (0 grid-current) grid-driven 
tetrode or pentode is simpler to build than two cathode-driven/g-g stages 
in series.   AB1grid-driven also has the advantage of not requiring a 
flock of tuned input circuits with a Q of 2 to 3 -- which naturally do 
not cover much frequency range.  

> I've been working with my latest project ( 
>still on paper ) which is a pair of 4-400As and it appears that all other 
>things being equal, about 200-250 watts more is available in AB2 operation 
>than is available in AB1. 

**  True enough, however, by increasing an AB1 stage's anode supply V by 
the amount of the screen-V (800V in this case), the anode-screen V will 
be the same as in AB2 cathode-driven/grounded-grid, so the output in AB1 
will be quite similar to that in AB2.  Thus, the output from a g-g/AB2 
config operating at 4000V will be c. the same as grid-driven AB1 config 
operating at 4800V.   For 4-400As in AB1/SSB, 4800V is do-able.  For AŘ 
or RTTY, forgetaboutit.

> I guess I understand why this is but don't 
>quite understand why AB2 grid-driven service gives poor linearity yet AB2 
>cathode-driven service is acceptable.  
>
>I decided to try grid driven service because it simplifies bandswitching ( 
>given the parts in my junkbox ) and in passive configuration, makes for 
>construction simplicity ( no tuned input needed ).
>
>Again.... many thanks to all and Happy Holidays
>
>RD


-  Rich..., 805.386.3734, www.vcnet.com/measures.