Hi John...
I refuse to be drawn into a discussion of what I was doing socially in
1947. Just figured it would have taken you at least 50 years to get so much
QRO experience! (Not kidding - but that's a compliment.)
When first discussing 3CX15,000B7 for MRI applications with Eimac guys in
San Carlos about 12 years ago, the about-to-retire chief engineer (or maybe
he was the just-retired chief engineer) told us they'd had substantial
problem with the tube failing prematurely in (sw broadcast, I think)
applications. Seems they finally determined that it would break into a
cavity-mode L-Band parasitic oscillation if idling Ip was great enough to,
I suppose, create high enough gm. Because the parasitic occurred entirely
inside the anode cavity, it did not couple significantly to the external
circuit. As I recall, that made it both extremely hard to discover and
identify, and virtually impossible to suppress other than by limiting Ipo.
Fortunately the critical level was modestly greater than what's required
for optimum linearity at 20-30 kW peak in pulse duty, and we've had
excellent service from the 3CPX5000A7, which is internally identical.
Think you've described the same phenomenon below, though with a larger
tube. Thought the above might be of interest in confirming your assessment.
(See, there's that evidence of your 50 years' experience again!)
73, Dick W0ID
>>> -----Original Message-----
From: John Lyles [SMTP:jtml@lanl.gov]
Sent: Wednesday, April 07, 1999 5:45 PM
To: amps@contesting.com
Subject: [AMPS] parasite
I see a parasite has invaded the ever quiet mailing list.
First, I have to comment on Dick's comment:
>And BTW, I also think those old ARRL Handbooks that John mentions, cc.
>1947-56 or so, covered the subject in a much more practical and helpful
way
>than more recent publications. Guess it's because lots of us were in high
>school back then, had no choice but to brew our own amps, and (blush!) had
>barely even heard of "Q."
Speak for yourself, OM! I wasn't even a glimmer in my Dad's eye in '47.
'56, well, I was crawling in diapers then. Hey, you guys must have been
something in high school. Building amplifiers. Didn't ya chase girls behind
the stadium and all that stuff?
---------------------------------------
RECENT EXPERIENCE WITH PARASITES
My 200 kW amplifier made of two independent out of phase amplifiers at 2.8
MHz, can be induced to take off, if I crank up the idling plate current
about about 50 Amperes. I do this by raising screen voltage above 1.5 KV
DC. Every time. Either tube does it. I have separate screen power for each
one. But I do not even need to operate at this high a quiescent current,
in class A. At 40 Amps, it is stable as a rock. Tuners can be moved all
over. When I spoke with the tube manufacturer, they only scrunched their
faces and admired me for pushing it this high. They hardly ever exceed 20
Amps of plate current during testing.
When driven to take off, it immediately crowbars the power supply, the
plate current goes high, and triggers the mercury ignitron to slam off the
HV very fast. It is impossible to see the frequency or mode of parasitic,
as it is too fast. Plus this is a pulsed machine, only 5% or less duty
factor, meaning it is pulsed at 20 Hz.
It doesn't bother me to know it can do this, as it is beyond the range of
operation that is needed for 100 kW per tetrode. With about 27 dB of gain,
these jugs are screaming as it is, in class A.
I know from cold measurements (GDO, network analyzer last summer) that the
dominant mode for Ctune and L resonance is about 40 MHz. This one doesn't
seem to be a problem throught the tubes. The isolation is good, the sockets
are exceptionally isolated. The other possibility is either a higher
frequency mode around 100 MHz, or a UHF thing, due to the circumference of
the region between plate to screen cylinders
inside these 80 lb tubes. Yes, they have eyehooks, not handles.
My parasitic suppressors were never installed, they work on the 40 MHz
resonance. If I get time, I may try them under power, with fingers crossed.
For academic reasons, and to show that copper straps and globars work too.
Best to just lay off the juice, and run the thing where it likes running,
dissipating about 30 KW in each tetrode for now. Good night now.
73
John
K5PRO
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|