>Hello Rich
>
>I seem to be having trouble posting to AMPS, so thought I'd better send
>you this. If you can send your reply to AMPS, that should get the
>discussion back on track.
>
RR, Ian
>Message begins:
>
>I tried to send this message yesterday...
>
>
>2 Wrote:
>
> >>>-----------------------------------------
> >>> "VARIAN ElMAC, 301 Industrial Way
> >>>San Carlos California 94070 1 U.S.A. / Tel. (415) 592-1221 TWX 910
> >>>376-4893
> >>>February 18, 1986
> >>>Your letter about parasitics is quite interesting, and it appears your
> >>>two tubes have had the same trouble. The emission was poor on test, and
> >>>consequently other test results looked bad. The tube engineer then cut
> >>>them both open for an internal examination.
> >>>Both have been badly overheated internally, the apparent result of an
> >>>oscillation condition. The grid in these tubes is gold plated and if
> >>>overheated the gold vaporizes off, of course, and some of it inevitably
> >>>lands on the oxide cathode, and that poisons emission.
> >>
> >>Thank you for the direct quote from EIMAC. From time to time it seems a
> >>good idea to separate what EIMAC said from your own opinions about what
> >>that implies.
> >
> >** In February, 1986, Eimac convinced me that the circumstantial
> >evidence I was observing was indeed due to an intermittent oscillation
> >condition.
>
>See below...
>
> >>
> >>The part about vaporization due to the tube having been "badly
> >>overheated internally" makes perfect sense, but it only says
> >>"apparently" due to an oscillation condition. EIMAC cannot say for sure,
> >>because EIMAC cannot know what happened - all they see is a tube that
> >>has been overheated *somehow*.
> >
> >** You need to see a sputtered grid for yourself with a 30x microscope,
> >Ian. The grounded end of the gold-plated grid often exhibits bare
> >patches of molybdneum base-metal. The other end of the grid exhibits
> >virtually no gold evaporation. To me, this seems similar to current
> >distribution in a quarter-wave vertical antenna. This is what leads me
> >to conclude that the current in a gold-sputtering episode is most
> >probably UHF.
>
>Sorry, but none of that theory ties-in with what VHF/UHF amplifier users
>know about the same tubes.
** Unloaded regeneration does occur in a UHF amplifier because the anode
is loaded at UHF. In an HF amplifier, this is not the case since the
Pi-network output section is a low pass filter.
>
>The grid "monopole" is heavily capacitively loaded by the surrounding
>anode, so the current distribution wouldn't be that of a free
>quarter-wave resonator - it would be much more uniform.
** So why isn't the evaporation pattern uniform?
>:>Also there is no
>sign of a grid resonance as a "monopole" up to at least the
>manufacturer's quoted maximum frequency for normal operation - which for
>the tubes we're talking about can be 200-400MHz. Above that frequency
>the gain drops off rapidly, so violent oscillation seems very unlikely.
>
** Eimac's 8877 development team came to a somewhat different
conclusion.
>I don't know why you observed what you did, but a quarter-wave resonance
>does not seem a likely cause.
>
** did you check this with a UHF dipmeter?
> >>
> >>Also EIMAC notably do *not* say anything about the frequencies involved.
> >>
> >** During our telephone conversation, I asked Mr. Foote about the
> >possible frequency of the "oscillation condition". He speculated that it
> >could be as high as UHF, but that the 8877 development team did not
> >measure a frequency.
>
>And weren't Mr Foote's spur-of-the-moment speculations later repudiated
>by the development team?
>
** No. When the advertiser's feces hit the fan over my article "The
Nearly Perfect Amplifier" [Jan 1994 *QST*], Tom Rauch's friend at Eimac,
Reid Brandon, the customer rep, told Paul Pagel ['Technical Topics'
Editor] at HQ that Mr. Foote was not authorized to release the
information on gold sputtering from the 8877 grid.
> >- The problem with a gold-sputtering event is that the window for freq.
> >measurement remains open for perhaps a few milliseconds until the gold
> >vapour cloud causes a +HV to grounded-grid flashover - which rather
> >rudely shuts the window. This is why I look for post-mortem clues with a
> >dipmeter, a saw and a 30x-microscope. Who know but that this is a
> >fanily trait since my cousin Virginia is a county coroner?
>
>Absolutely... but your cousin is probably more cautious about
>speculating beyond the actual evidence!
** Eimac was the first to see the evidence and make a conclusion. The
autopsy evidence I saw myself convinced me they were on the right track.
>Given the difficulties of
>post-mortems, "open verdict" still seems about as far as we can go.
>
** "We" = Ian, Tom, Reid, Dick E., and no doubt others.
>
> >- Alpha currently admits that gold-sputtering is a problem, but that
> >it's really and truly caused by the (idiot) consumer/operator driving the
> >amplifier with too much HF power.
>
>It's more likely a combination of several things. Very few users
>actually know what their grid dissipation is (it is not simply V_bias *
>Ig1). Many users cannot credit that a few seconds of overload while
>tuning-up can do the damage.
** At what temperature does gold boil? What is the approximate mass of
an 8877's grid
> Many older amps do not have protection
>against excess grid current. And also the tube manufacturer has a
>problem: they have to decide on a grid dissipation rating very early in
>the life of the product, and if long-term experience indicates that
>rating was optimistically high, they would then find it very hard to
>revise the rating downward.
>
> >On this side of the pond, such is
> >typically called "Not Invented Here Syndrome".
>
>"Mutual Denial", more like.
>
** same animal
> >- One of the reasons that my October, 1988 article on parasites was
> >published in *QST* was because a plethora of kaput TL-922 owners had been
> >kvetching to Newington/HQ that Trio-Kenwood "Service" had been telling
> >them that their 922s were damaged was because they were (stupidly)
> >"bandswitching their amplifiers while transmitting at full power". HQ
> >had apparently heard enough.
>
>There was clearly a genuine problem with the VHF stability of that amp
>(not to mention T-K's approach to customer service). You and ARRL both
>did well to confront and solve it.
>
>But not all amps are TL-922s...
** who claimed they were?
>
> >- Most manufacturers loathe admitting mistakes and fixing problems.
>
>Certainly - but when a problem may be due to a complex combination of
>the amplifier, the design of the amp and/or the way the customer uses
>it, it can be hard to judge who should admit to how much share of the
>responsibility. It's wrong to blame any party automatically.
>
** The blame is affixed by evidence. There is no reason for a 3-500Z's
filament to be forced sideways by a customer unless he/she is Ham
acrobatic airplane pilot who likes to operate aero-mobile whilst practing
aerobatic manouvres. Autopsy evidence shows that the filament can be
bent sideways in normal earthbound operation.
>And it's *very* wrong to transfer the same judgement from one case to
>another!
>
** The evidence speaks for itself.
- R. L. Measures, a.k.a. Rich..., 805.386.3734,AG6K,
www.vcnet.com/measures.
end
|