> Jon, you stated previously that you were in the resistor business.
> Why don't you provide a mini lecture on the inductance in helical wound
> "metal film" resistors of the type that are probably going to be
> encountered in the 5-10 Watt class on RF suppressors? Here in the third
> world, even 1/4 Watt "metal film" resistors turn out to be helical wound
> jobs, readily proven by burning the paint off them and inspecting the
> remains... Surely if the function of a resistor is to emulate R only, then
> inherent inductance must be kept low, and carbon composition types are the
> preferred option?
Or you can put a series capacitor in series with the R and cancel
the reactance at the frequency of instability.
>
> Another thing that worries me about this parasitics debate:
> 1. If certain tubes have a tendency to take off at low VHF freqs, ~110
> MHz, what about 28 x 4 = 112 MHz, 21 x 5 = 105 MHz, 14 x 8 = 112 MHz, etc?
> The point is, if there is a tendency to oscillation out of band in a
> linear, a sniff of unwanted harmonics would drive it over the edge.
Oscillators don't work that way. They either have enough feedback
to overcome loss and they oscillate, or they don't.
Now you can vary the loss or gain by adjusting controls or
voltages. You might have a loose connection that, when you
smack the thing, a connection breaks and the oscillation starts.
You can have heat related gain or loss changes, but once
regeneration exceeds loss the thing oscillates. If loss exceeds
regeneration, it won't.
"Banging" it with a harmonic is certainly not any requirement to
start an oscillation, especially since a typical HF PA has a shorter
than 360 degree conduction angle and is thus rich in harmonics all
by itself even with pure sinewave drive.
This
> could explain why there are thousands of unmodified, stock SB 220's out
> there, which have never hiccupped for years - the drive has low harmonics.
No, they don't "hiccup" because they don't have gassy tube, or
because the builder tunes them correctly, or because the exciter
doesn't have 3x normal power in leading-edge transients that
overdrive the PA and cause excessive RF voltages in the tank.
Sometimes a bad antenna connection will trigger an arc.
One thing we can be sure of, the LAST thing that will trigger an arc
is a harmonic or any frequency far above the operating frequency.
If we have an ounce of knowledge about components and networks
we know the impedance looking across the plate capacitor is less
on the second harmonic and higher, and so any voltages developed
there will be less as frequency is increased...not more.
By an overwhelming amount the largest problem with arcing occurs
because of fundamental frequency voltages, not higher frequencies.
> 2. Another question, does the input circuit on a linear such as the SB 220
> have sufficient low pass attenuation to prevent direct coupling into the
> tube? Most input circuits are tuned for max power output, minimum SWR as
> seen by the exciter, with nary a thought given to harmonic attenuation,
> anyway. Never mind RF feeding into the input from poor local conditions or
> high SWR on the output line and so on.
All input circuits in non-class "A" PA's should present a very low
impedance at harmonics of the operating frequency from the
cathode to ground in a cathode driven PA. That's because the
conduction angle of the amplifying device generates harmonics,
and if the input circuit presents a high impedance to even order
harmonics the switching-state waveshape softens and the tube
dissipates a lot more power for a given output power.
If you insert a second harmonic resonator in series with the
cathode in a 180-200 degree conduction PA, efficiency can almost
halve!
The tuned input is not an issue except for matching, and the
impedance presented to the cathode-anode current at even-order
harmonics.
Since the amplifying device itself generates harmonics, and since
oscillators can't "hang" in a state where the regeneration exceeds
loss with oscillating (even without any external influence), input Q
and exciter purity is a non-issue except for the exceptions I
outlined above.
73, Tom W8JI
w8ji@contesting.com
--
FAQ on WWW: http://www.contesting.com/ampsfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|