What we really want is a maximum enrgy transfer into a
dissipative resistance at the frequency where a VHF
oscillation might occur, and minimal transfer at other
frequencies.
This means the inductor path has to be an almost pure
reactance, while the resistance path has to be a low
impedance compared to the inductance at VHF.
When we have only two pure components involved, and inductor
without resistance and a resistor without reactance, the
slope is 1:1. Say we have a reactance of 10 ohms at 30 MHz
and a resistance of 100 ohms. Every time frequency doubles
the reactance doubles. The resistance is constant.
If we make the inductor resistive we actually do the
OPPOSITE of what is normally desired. Now the slope is
less, so for the same dissipation at 30 MHZ we have a LOWER
impedance at higher frequencies.
The correct solution would be to add a capacitor in series
with the resistor, so the shift is greater as frequency is
increased. This means for the same dissipation at 30MHz we
can have much more resistance at VHF.
The optimum resistance (or impedance) of the suppressor
depends entirely on the anode to tuning capacitor path
impedance. When that path has a lot of reactance, such as
one with long thin leads, we need MORE resistance to load
the system. This means a suppressor with more inductance and
higher resistance. The problem is this can cause heating at
ten meters. The solution is really to decrease the impedance
of the path from anode to the chassis at VHF by using wider
and shorter anode leads.
The grid is also important. The grid path determines the
frequency of instability more than any other single
parameter. The grid leads must be as short, thick, and
direct as possible. The actually oscillation mechanisim at
VHF is almost always as a TPTG oscillator , and so when the
grid is moved UP in frequency the oscillation, if there is
one, occurs higher in frequency.
It really has little to do with "VHF gain" of a tube. As a
matter of fact the most stable tubes are those that have the
highest VHF gain when used in amplifiers. 8877's are
significantly more stable than 833's for example.
As for arcing, the idea a VHF parasitic causes an arc in an
otherwise healty tube is largely nonsense. The peak anode
voltage compared to the voltage breakdown is the problem. An
oscillator really doesn't develop any more voltage than an
amplifier when driven to the same grid current into the same
load impedance. If you have a healthy 3-500 it will arc at
12kV or more peak anode voltage, and neither a parasitic nor
a desired signal will reach that level before another
component fails. If the tube is bad, say it fails at 6kV, it
can arc from anything that produces 6kV of peak anode
voltage....and that can be anything from a desired signal in
normal operation to an oscillation.
It's very poor thinking to assume decreasing the slope of
resistance with frequency (by lowering inductor Q) somehow
improves the suppression at VHF for a given dissipation at
HF. That idea shouldn't make sense to anyone who thinks
about how the system works.
If the problem is not enough resistance at VHF then the
solution is to increase inductance at HF along with the size
of the shunt resistance. This would keep dissipation the
same at HF while increasing loading at VHF by making the
series resistance a larger portion of the impedance of the
anode path. If that doesn't work well enough than a
capacitor can be placed in series with the resistance or in
shunt with the inductance to increase the slope of
resistance to reactance change as frequency is increased.
The important thing to understand is the entire system and
how it works. The grid needs a short direct ground to the
chassis, the anode path needs to be short with low VHF
impedance, and the suppressor must dominate that path at VHF
and look nearly invisible at HF if you want to maintain ten
meter performance.
The idea a VHF oscillation will make a healthy tube arc is
almost silly. It's grasping at very unlikely straws.
73 Tom
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|