[Amps] A Meeting Ground

[R.Measures]

On Nov 5, 2004, at 1:29 AM, Steve Thompson wrote:

> On Thursday 04 November 2004 20:14, Dennis12Amplify at aol.com wrote:
>> Bill,
>>
>>  I have been reading your posts lately and want to commend you for  
>> your
>> 'sage' advice.
>>
>>  I also enjoy Rich's posts but lately he seems to be a bit  cranky.
>>
>>  I guess I would too if I kept having to fight the parasitic  battle.
>>
>>  I kept thinking about the phase angle thing and realized that for a
>> grounded grid amp with .12pfd of plate to cathode capacitance to 
>> oscillate
>> at  1000 megacycles, the uhf circuit gain at that frequency would 
>> have to
>> be greater than unity at a point where there was either no phase 
>> shift, or
>> -/+ 360  degrees of phase shift between cathode and anode.
>>
>>  With the feedback impedance of 1326 ohms at -90 degrees voltage  
>> dividing
>> with the input impedance of the tube (another vector but a low  
>> impedance
>> one), It would take a significant amount of stage gain to cause
>> oscillations to occur.
>
> Sounds right to me. By 1000 MHz I suspect that the phase angle of the 
> internal
> feedback might not be exactly -90 degrees - but I also doubt that it's 
> zero.
>>
>>  And wouldn't there also be an additional capacitive voltage divider
>> between the plate to cathode capacitance and the grid to cathode
>> capacitance  further reducing amplitude of the feedback signal?
>
> Yes. Consider what happens in a hf amp if the cathode circuit provides 
> a low
> impedance path to ground at vhf.

Yo, Steve --  It would seem that with a low-Z path to gnd 
(circuit-common), more of the VHF feedback signal (result of delta-I 
ringing of the anode's VHF resonant circuit) would be shunted to gnd, 
and less of the VHF feedback signal would be arriving at the 
cathode/input -- which would obviously be a serious error if one is 
designing a VHF oscillator.
-  In the SB-220, there is a length of RG-58 coax between the tuned 
input Pi-networks and the cathodes of the 3-500Zs.   A dipmeter reveals 
that this section of coax creates a VHF resonance that is close to the 
frequency of the ringing in the anode's resonant circuit.  This 
resonance keeps the ringing signal from being shunted to gnd, and 
allows more feedback signal to go into the cathode.  Essentially the 
same situation exists in the TL-922, except that the ringing frequency 
is c. 10MHz higher and the RG-58 is a bit shorter.  Both amplifiers 
have a track record of big-bangs, toasted VHF parasitic-suppressors, 
blown grid chokes, , toasted 10-15 m contacts on bandswitches, and 
grid-filament shorts.
--------  Editorial:  Tube manufacturers list feedback C in pF, which 
are units of 10^–12 Farads.  For instance, a 3-500Z has 0.15pF, or 0.15 
x 10^–12 Farads of C.  Since feedback-C is so small, most engineers 
seem to ignore it.  If feedback-C was listed in femto-Farads (10^–15) 
F, the feedback spec on a 3-500Z would be: 150fF -- and, who knows, 
perhaps feedback C might get more respect?

cheerz

...
Richard L. Measures, AG6K, 805.386.3734.  www.somis.org