[AMPS] UPDATE ON 3-1000Z OSCILLATION

[Tom Rauch]

> from the output.  I remember the shielding used to isolate the input
> from output was much greater than most present day linear amplifiers. 
> First hand experience taught me it was better to design it right in
> the first place from an isolation standpoint than to retro fit it with
> shields made from old tin cans.

Gain in a 6AK5 grounded cathode amplifier is much higher than a 
typical GG amplifier. Isolation of input and output is more important 
when gain is higher.
 
> It is tempting to tell ourselves that the amplifier is grounded grid
> and bypass good construction techniques because feedback is less of a
> problem. The fact that parasitic oscillations are still with us is
> proof that grounded grid is not a cure all or excuse for sloppy design
> and construction.

The problems with oscillations in a power amplifier are virtually all 
rooted in the physically large tube, and mostly center around the 
grid structure inside the tube and the length of leads to get outside 
the envelope of the tube to the chassis. 

A high-power amplifier is clearly much different than a high-gain 
small-signal device. 
 
> Uhf amplifiers rarely have parasitic problems, yet use no parasitic
> suppressors.  That should tell us something about good design.  We
> should probably review a few uhf designs before building  our next
> amplifier and save a few headaches.

The most stable tubes are tubes that are useful at high 
frequencies, assuming grid leads are kept short. The most difficult 
tubes to stabilize are almost always tubes that have low maximum 
frequencies of operation.

That's why the 8877 is so easy to stabilize, and tubes like 811A's 
or 3-1000Z's are more difficult. If we look carefully, we see it is 
tubes with long thin grid leads that are most troublesome. They are 
often unstable no matter what the layout, because at some 
frequency the grid has a very high impedance to ground no matter 
how we ground the grid.

In that case, we need to make sure the anode (or some other 
element) does not have a similar resonance.

At HF...it takes a very poor layout for external feedback to be an 
issue, because most of the feedback is directly from anode to grid. 
The most important thing is grounding the grid as best as possible, 
and keeping the length of the anode connection short so we can 
control the impedance in the anode easily.


73, Tom W8JI
W8JI@contesting.com 

--
FAQ on WWW:               http://www.contesting.com/FAQ/amps
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com