On Jun 27, 2006, at 8:08 PM, Tom W8JI wrote:
> From: "R L Measures" <firstname.lastname@example.org>
>> As I see it, no matter what kind of modulation one is
>> trying to amplify, at the point when an incremental
>> increase in input no longer increases output (which call
>> the transfer function) by the same factor as was observed
>> at lesser currents, the cathode is running out steam and
>> the tube is beginning to lose linearity.
> Not true. Even a curved line does not necessarily produce
> IM3 or IM5. (harmful IM3 is 2 * F1 - F2, or 2*F2-F1, IM5 is
> 3*F1-2*F2, 3*F2-2*F1 and so on through all odd-order
RE: A curved line: Since the slope of the line is constantly
changing, it means that the transfer characteristic is too, and that
means it's making most all types of doo-doo.
> There were two posts to this reflector, one from
> Steve Katz and another from Doug Coffman, indicating they
> measured a linear transfer response when the load line was
> set over 550mA during initial tuning. You might look back at
> the recent archives.
No sale. Changing slope tells the tale.
>> Even though there is a 3a line on the 3-500Z/ 8002
>> characteristic curves, does this mean that all will be
>> well in linear service at 3 peak anode amperes?
> Comparing one set of operating parameters in one system to
> another entirely different set in another system doesn't
All of the constant curves use the same system.
>> If this were the case, at 4kV, single 3-500 linear
>> amplifiers could produce 2500w PEP.
> What approximation did you use?
I knew that an 8877's constant current curve for 3a-pk (c. 1A avg) is
fairly flat, and that with 4000v @ 1A in it would do c. 2500w out
with total IMD under 40db down.
>> OTOH, Tom may be right in saying that tuning up at 550mA
>> produces good linearity in an AL-80 or SB-1000, but from
>> the impure matter I hear from these two amps on the air
>> that are driven with 100w, I have doubts.
> Interesting. You formed your conclusions because (over the
> years you have been listening) you "heard" two wide signals,
> one from each model.
Not true. From a plethora of on the air observations of both
models. A friend bought one, it produced feculence with 100w drive,
he added 25-ohms of cathode RF-NFB, tuned it up at 420mA instead of
550mA, and that apparently cleaned it up.
> The test conditions were:
> 1.) unknown quality communications receiver at your end
Can he be serious?
> 2.) unknown measurement signal-to-noise ratio at your end
> 3.) unknown exciter IM quality at their end
The exciter proved not to be the culprit since the signal cleaned up
with the cathode RF-NFB R, so the following straws are neutralized.
> 4.) unknown operation of the offending radios and amplifiers
> at their end
> 5.) unknown condition of the offending radios, amplifiers,
> and antenna systems
> I have a room full of newer calibrated measurement equipment
> including equipment that can automatically measure occupied
> I live in a rural area with nearly a hundred acres of
> directional receiving antennas so the antennas can focus on
> a desired signal as clearly as possible.
> With all of that at my disposal I couldn't tell for sure
> over the air if it was the exciter, operator, amplifier, or
> even something arcing or non-linear in their antenna. I
> would need to be right in the room with the offending gear
> and test equipment to tell anything about the actual cause.
> I can't think of anyone else who, over the air, can tell if
> it is the amplifier, the exciter, the antenna, a feedline
> connection, or the operator with any certainty. I also don't
> know many people who consider a sample of two bad signals
> under such poor test conditions justification for going on a
> campaign against someone or something.
The A - B test with and without the NFB R told.
>> I don't see how because 550mA corresponds to c. 1650mA
>> peak, which is well into the non-linear region.
> I'm not sure where your numbers come from, since you never
> really say how you get them.
I open my eyes and look at the constant current curves.
> 73 Tom
> Amps mailing list
R L MEASURES, AG6K. 805-386-3734
Amps mailing list