[AMPS] RE: 8877 failures; thoriated tungsten vs. indirectly heated tubes

[richard w. ehrhorn]

In addition to Tom's comments, some additional inputs. Welcome to the history & mystery channel...

The seminal theory and discussion of thoriated tungsten emitters was published sometime in the late 20's - by an RCA guy I think. The basic R&D and tooling is long since amortized! Indirectly-heated cathodes for tubes operating above about 2.5 kV B+ didn't come along until after the end of WWII so far as I know. 

Modern high-linearity, high gain cathode-type tubes including the 8874, 3CX800A7, 8877/3CX1500A7, and 3CX15,000B7 have "striped cathodes," picket-fence style grids with conductors precisely aligned to the cathode stripes to reduce electron interception and improve linearity, and very close spacing between the cathode cup and grid - on the order of 20 mils in the 8877. Oxides used to coat the cathodes emit electrons via a complex process, comprehension of which requires a better understanding of solid state physics than mine. The oxide preparation and application process seems to be partly science and partly black magic - and susceptible to contamination not unlike some semiconductor processes.

Around 1988 the manufacturer attempted to improve the 8877's cathode emission without changing heater ratings. They changed the metal used for the "cathode dam" - a very thin cylinder designed to support the cathode while conducting away very minimal  heat (to keep required heater power down). Unfortunately they evidently didn't check thermal expansion properties of the new dam material, resulting in a big mismatch between it and the cathode cup to which it was welded. Result: at each heater turn-on the cathode dam was forced outward by the expanding cup, and it didn't contract fully to its previous diameter during cool-down. So it ratcheted radially outward toward the grid (nominally ~0.020" away) by a fraction of a mil over each heater on-off cycle. For six months to a year, evidently, 8877's in all kinds of service (ham, MRI, etc.) failed enmass with grid-cathode shorts - typically after about 200 hours of service... plenty of time to complete in-plant testing and burn-in, installation in the field, and probably up to a year or more of amateur service before the problem showed up. By the time we & our customers recognized the failure pattern, we had at least 500 8877s of that "version" in field service. And for nearly a year the only thing we could get to replace them was more tubes with the same defect, doomed to fail at ~1/100th normal life. I guess nearly all circa 1988 8877s failed this way - we know they did in MRI!

To make it worse all around, the initial tube diagnosis and "fix" took around 6 months and didn't improve the problem. Another 6 months or so passed before good tubes started to ship again. Nothing similar has occurred since and to the best of our knowledge the Eimac 8877 has been a model of reliability. If not abused, 15,000-20,000 hrs is typical life. But that year-long-plus episode cost at least one company its biggest customer and literally millions of dollars.

That's why W1AW (and anyone else) had problems with premature 8877 failures due to shorted grids around 1988-92. 'Course we weren't making ALPHA 77s at the time. And all this happened years before the new owner, CPI, took over Eimac so it might even be news to them.

Can anybody shed more factual (maybe "inside") light on the 8877 cathode dam episode? Some of us would love to know more.

73,  Dick W0ID



> From:          Peter Chadwick <Peter.Chadwick@gpsemi.com>

> At what emission level does an indirectly heated cathode become an
> impractical emitter?

It's mostly a function of surface area with indirectly heated tubes, 
so they can have very high peak emissions. There are no hard and fast 
rules for indirectly heated tubes.

They are permanently damaged (they lose emission) by either excessive 
high voltage, or excess peak emission demands. Running low filament 
voltage, operating the tube without enough warm up time, or drawing 
too much current will ruin the emission.
  
> Or in other words, why thoriated tungsten for tubes
> like a 3-500Z?

Thoriated tungsten tubes saturate at about 125 mA or so per watt of 
filament power. No matter what you do with other voltages, you can't 
get more than that out of the cathode.

> So why thoriated? Is it
> much cheaper or what?

It is cheaper AND the tube is more forgiving. The grids are not gold 
plated, so that source of trouble is gone. You can often heat the 
grids until they glow without grid damage. The tube is not sensitive 
to warm up time, you can apply full power and drive even before the 
filament is hot without damage.

Not only are warranty expenses reduced, the production costs are 
less.

73, Tom W8JI

--
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