Tom,
What I was saying was that Bob was claiming that one switch worked at 3000 V and the other at 6000 Volts.
My statement was that his comment and thinly veiled insult about running the "6000 Volt" switch at 3000 Volts doesn't make sense. The fact is Bob was wrong about the voltage requirements on the B&W850/850A and he was dead wrong about what I had done and what I had claimed. I was just stating here that you can obviously run a switch at a lower voltage that what it is rated at. So even if what Bob claimed I did was true, it wouldn't matter.
I agree with you that Q and design impedance vary with frequency. If you read my previously long message, you would see I have gone through a pretty detailed mathematical analysis on the subject.
I was just calling Bob's insult what it was - specious.
73,
Jon
NA9D
W8JI@contesting.com wrote:
> > Since the plate impedances the 850A and 851 (I know not about the 852)
> are designed for are 2000 to 4000 Ohms, it matters not at which plate
> voltage you use them except for the fact that the 851 is designed to
> be operated below 2000 VDC at 250 mA for SSB and CW.
>
> So your statement appears to be specious at best and doesn't
> completely make sense.
I'm afraid it does make sense Jon.
The is no "design impedance" for a switched inductor sitting on
your table, because the design impedance varies with loaded Q
and frequency.
You could say the "design impedance is 2500 ohms at 3.7 MHz
with a loaded Q of 12", but that very same inductor would have a
design impedance of about (but not quite exactly) 5000 ohm if the
loaded Q was set to 6, or about 1250 ohms if the Q was set to 24.
The operating Q can change all over the place just by changing
power levels and capacitor settings, as can the impedance.
People really get far too carried away with Q, because efficiency
changes very little with Q as long as tank losses are low. Factually
the lower the loaded or operating Q becomes, the higher the
efficiency becomes until Q is the square root of the impedance
ratios plus a tiny bit. When loaded Q drops below that value,
network phase shift is 90 degrees or less and the network no
longer behaves like a Pi Network.
The tank voltage can be many times the dc anode voltage, if the
tank is "unloaded" or "underloaded", depending on the conduction
angle of the tube and the impedance presented to the anode. But
in a properly tuned class AB PA the peak voltage in the tank is
always a bit less than the dc anode voltage.
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