Rich wrote:
> >Look at it this way. Suppose you did have a "parasitic". Why would
> >the parasitic make the tube arc?
>
> It doesn't. The big-bang arc is from the B+ circuitry to gnd. The arc
> leaves no arc marks inside a shorted tube.
Most tube faults are gas related. Any arc from anode to grid or
anode will cause a big bang. It does not have to be from the HV
circuitry to the chassis EXTERNAL to the tube.
> >When you get a big bang like that, it is because the HV is faulted
> >to "ground" through a low resistance non-current limited path. If
> >that low resistance path is through a normal tube, current can't ever
> >reach more than 20 amperes.
>
> The tube is not in series with the B+ to gnd arc path.
Then the GG amplifier might not work well.
The anode should be connected to the HV, the grid directly or
indirectly to the chassis (which connects to B-), and the filament
connects through the diodes that failed to the B-.
> Why don't grid/filament shorted tubes show gas after they are removed for
> high-potting?
Because a grid-filament short is a geometry problem. A HV fault is
gas or barnacle related. The two are unrelated, unless a HV arc
happens to damage a grid. That is an unlikely scenario in a 3-500Z.
> >The SB-220 sets itself up for problems by not directly grounding
> >the grids. If the grids were directly grounded, any fault path through
> >the tube would be less likely to involve the filament because the grid
> >would be a "shield".
>
> A directly grounded 3-500Z grid is resonant c. 80MHz. Above that freq.,
> the grid does Not act as a shield. SB-220s oscillate at approx. 110MHz.
> The 3-500Z is rated at 110MHz max. for Oscillator and Amplifier service.
No, that isn't true. It is a commonly propagated myth, but totally
incorrect.
The frequency at which the grid appears to be "floating" in a 3-500Z
is near 200 MHz, unless the grid leads to the socket are long. I
have measured that on my network analyzer.
Adding the capacitors in the SB-220 does not raise the frequency
where the grid is parallel-resonant, it lowers it. The capacitor puts a
single large dip in the feedthrough, but at the same time
deteriorates shielding on all other frequencies.
It was a bad idea, copied from the Collins 30S1 at the suggestion
of Bill Orr. Bill did not understand that the screen in the 30S1
shields the input from the output, and floating the control grid of the
tetrode made little difference. Copying that circuit to a multiband
GG amplifier is poor engineering.
I'd be happy to supply photographs of a 3-500Z in a test fixture,
with S21 attenuation being measured on a Network analyzer with
sweep from 1 to 1000 MHz, both with and without capacitors.
> >You can still have excessive current out
> >the cathode if the tube is gassy, but at least you have some
> >reasonable chance of the current being limited.
>
> ? Say what?
What.
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
|