[AMPS] L band oscillation inside tube

[John T. M. Lyles]

Dick,

>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.)

Thanks, but ham radio was the source. Plus starting with MW broadcasting as
a teenager, around a Gates BC500T (500 Watts, 1.25 MHz) in South Carolina.
Lets say it was more like 29 years.

We had a safety class at work last week, on RF. It seems that nonionizing
radiation has caught the attention of the safety experts, who normally deal
with 60 Hz powerline hazards for electricians, or deal with alpha, beta,
gamma ionizing radiation. So this expert starts talking about RF burns, and
the phenominolgy of RF current, skin effect, how it compares to power mains
frequency. Smarty pants here (me)  raises his hand and tells the class
about the numerous RF burns I have gotten over the years, and how long they
take to heal. Those little white marks on the fingertips, which smell bad,
and hurt when the arc occurs as you touch something. I get very few
anymore, since the amplifiers and feeders are too dangerous to be poking
around while energized. The teacher then says he has never seen or gotten
an RF burn, but heard stories. And everyone looks at me as if I was some
sort of daredevil freak who walks barefoot over broken glass and things
like that - doing shots of tequila while drving down the freeway, you know.
Jeez, where have all the real RF people gone? A bunch of bed-wetting
computer geebs that have never experienced that lovely aroma of burning
flesh, or melting G10, or that tinking sound in a tube amplifier after the
power is shut off? Gads, I must be really off in a special corner of
technology. Anyone else feel this way?

>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.

Great! I will add your experience to my 'list' (mental). I think those
internal parasitics do a lot of damage to tubes like that, as there is no
output coupling to load them. The Burle 7835 super power triode has a great
propensity to want to do this, when undriven while the plate voltage is on.
No one biases it with DC on the grid or cathode, instead opting for self
biasing using a DC resistance of about 0.25 Ohms. The cathode current is on
the order of 400 Amperes DC when driven, or less when undriven. Not a zero
bias triode by any means.

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