Hi Carl,
> Please explain the relationship between the resistor value and the coil
> Tom.
The rule of thumb in a tube with poor grid grounding is the
resistance has to nearly equal or slightly exceed the impedance of
the anode to chassis path at the oscillation frequency. Sufficient
parallel resistance in the suppressor insures a SYSTEM Q of near
unity, and sufficient damping to prevent oscillation.
The coil has to have enough reactance to allow the resistance to
dominate or nearly dominate the system at the self-oscillation
frequency. The suppressor must appear externally mostly as a
resistance.
Long thin anode leads and unstable tube types always dictate more
resistance be used. Higher resistor values, required by higher anode
path impedance and a less stable tube, always require larger shunt
inductances in the suppressor system.
In extreme cases the inductor can be parallel tuned, or the path
through the resistor series tuned. As the tube becomes more stable,
less loss is required. The suppression system is less critical.
The oscillation frequency and tendency towards instability (dictating
the amount of loss required) is largely controlled by the grid. That
is why tubes like the 811 and 572 use many turns on the suppressor
(and the suppressor design, as you pointed out, is critical) while
tubes like the 8877 often need no suppressor at all. "In between"
tubes like the 3-500Z use in between values in the suppression
system.
Treat the amp like you are building a VHF amp, especially the grid
and anode, and it will be much easier to stabilize.
> In commercial amps there appears to be no set pattern; but yet I
> assume that real engineers were involved.
I'm sure many amateur manufacturers use the Edisonian method, cut and
try. Many designs also copy each other, the Japanese did copies of
the SB-200 and 220, mistakes and all.
> >Series resistive losses (such as winding the coil from a lossy
> >material) in the coil have greater effect as the frequency is
> >lowered, while parallel resistances (the kind most people use) have
> >less effect at low frequencies.
> There I must disagree. Pure resistive losses due to skin effect become
> more pronounced as frequency is raised, not lowered. The same reasoning
> that is used in wire sizing for the tank coil should follow thru from
> anode to the output connector.
I'm not sure what you are saying. Are you saying the suppressor
inductor should be wound with the same size conductor appearing in
the anode path? If so, I agree. That is a good rule of thumb.
> So why are there no articles? Even "RF Design" in its heyday didn't
> broach the subject. Or any tube manufacturer has surfaced with a
> definitive paper or article.
Eimac has published some tech bulletins, but the subject has
mostly been ignored. I'm not aware of any detailed comprehensive
article on the causes and cures of instability in amateur
publications. I have no idea why there has not been one published,
but I do know tube design is a dying art.
> It seems that even those with formal training can confuse and cloud the
> issue. If one looks at it in the "black box" context your argument makes
> sense. If you climb into the box then Rs becomes a real part and is
> frequency dependent. One cannot continue to explain everything as a
> theoretical lossless circuit at RF; it is not as easy as basic 60Hz AC
> Theory. Even us uneducated types can understand that Tom.
I didn't mean to imply anyone couldn't understand it, or that anyone
is better than anyone else. My point is use of terms that can mean
many different things just confuses people trying to learn.
73, Tom W8JI
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|