[Johnson] Valiant Chirp - longish post

[Al Parker]

Hi Art, Chris, et all, again,
    Tom, W8JI, on another list, analyzed the problem.  Thought it'd be of 
interest here/now.  Perhaps you, Art or Chris, can try just grounding the 
cap more solidly before doing my bandaid fix, and let us know so whover has 
the problem next yr can benefit ;-)
73,
Al
=================
----- Original Message ----- 
From: "Tom Rauch" <w8ji at contesting.com>
To: "Old Tube Radios" <boatanchors at theporch.com>
Sent: Wednesday, December 05, 2007 8:18 PM
Subject: Re: Valiant chirp - fixed (long) rev.


>> I remain preoccupied with the fact that the capacitor was insulated from
>> the chassis yet grounded to the chassis.  Were they trying to avoid the
>> very circulating currents that they created?
>
> Whatever the root cause, the designer intentionally isolated the frame 
> (rotor) and brought it down to the area of the loading cap with a small 
> buss wire. It wasn't uncommon to do things like that just from habit.
>
> In all the PA's I've ever designed or looked at, probably  a few dozen low 
> power HF tube applications in the lot, harmonic suppression and stability 
> were always improved with short direct to chassis grounding of the tuning 
> cap. This is because the excitation path in single-ended designs is 
> through the time-varying cathode to anode resistance of the tube.
>
> http://www.w8ji.com/Vacuum_tube_amps.htm
>
> There are also other paths creating even larger rotor to chassis currents. 
> The loading cap itself is grounded to the chassis, and the return path for 
> circulating currents is through that cap. So you have tank Q and 
> circulating currents flowing through that cap.
>
> When they put a small skinny long buss wire from the rotor (frame) of that 
> tuning cap to the chassis, the frame of that capacitor lifts appreciably 
> in RF voltage. That voltage is what drives or diverts tank and PA currents 
> down the tuning shaft. It isn't the tank coil via induction, it is the 
> direct wired connection to the poorly grounded capacitor frame (rotor) 
> that causes the problem. While it might have been habit from other 
> designs, like push pull amps or breadboard amps, in a metal chassis 
> single-ended amp, floating the tuning cap frame cannot ever be the best 
> way to arrange things.
>
> Perhaps when the rigs were new, the shaft bushings further up that shaft 
> kept it grounded. The VFO shaft bushings were also probably clean and 
> shiny. Maybe that hid the problem.
>
> Whatever the case, the problem absolutely is not instability in any 
> external stages. The floating also didn't make the PA better. The PA stage 
> got better, not worse, when I replaced the upper insulated washers with 
> stainless steel washers (stainless because they will not corrode). 
> Harmonic suppression improved also.
>
> It's a very simple basic problem. There are a few amperes of RF current 
> flowing through the plate tuning cap. That cap is only grounded through a 
> long skinny buss wire. This means the rotor, since the thing was first 
> fired up, always had a voltage difference between the cap frame and the 
> chassis. That voltage would increase with frequency because the impedance 
> to ground would increase with frequency. On 40 meters the front panel has 
> enough current flowing to lift the shaft of the VFO away from the ground 
> potential of the VFO box. That directly couples the PA output back into 
> the VFO grid, and it pulls the oscillator.
>
> Since there is no reason to have RF currents flowing out that shaft, I 
> grounded the frame (rotor) of the tuning cap with metal contact directly 
> to the chassis with zero lead length by removing fiber washers and 
> replacing them with stainless. This is the way any single ended tank 
> should be constructed, unless it is built on a wood chassis.
>
> 73 Tom

 and my response:
12/05/07
Hi Tom,
    Thanks for clearing my thinking about this.  You can be sure that the
next time I have this beast out of the cabinet I'll be properly grounding
that capacitor, with SS washers.
    Thanks again for your help.  Your solution, and the other guys who
recognised the problem, sure saved me some time and head scratching.
73,
Al, W8UT
===============