[ARC5] Filaments - How Low Can You Go?

[Roy Morgan]

Richard K wrote:
>     I think it depends on the type of filament.  There are
> three generally used:
> Pure tungsten
> Thoriated tungsten
> indirect heaters using coated surfaces

ARC-5 folks:

I was quite interested to scan a technical article from the  Bell  
System Technical Journal about the development and performance of  
tubes developed by the early 50's for use in coaxial cable long line  
telephone systems (the L3), in particular the WE-437A.  The 437A is a  
miniature triode in 9-pin base that has  "high Figure of Merit" and  
uses an indirectly heated cathode with a coated surface.   Oddly, the  
exhaust tip is on the bottom between the pins. "Figure of Merit is a  
direct measure of the bandwidth over which the required amplification  
can be obtained."  The article is at:
<http://www.alcatel-lucent.com/bstj/vol30-1951/articles/bstj30-4-1103.pdf 
 >
and I recommend it for a view of a truly competent and professional  
approach on how high reliability tubes were developed.

The 437A cathode thickness is the same as the 6AK5 (0.0015", but  the  
cathode to grid distance is much less) and the expected life of the  
tube is some 15,000 hours.  That's a little less than two years  
continuous operation.  We have heard reports of R-390A's being run  
24/7 for many more years than that with still good performance.  The  
6280/416B is a planar grid amplifier tube (lighthouse shaped) meant  
for use at 4000 mc.  Its grid spacing  and cathode coating thickness  
are both much smaller than the 437A.  Without seeing it mentioned, I  
would think that the 416B expected life is equal to or greater than  
that of the 437A family.  (The 416B is gold plated.)

Further comments:
- Some poking around among WE data sheets tells us that the filament  
supply for their tubes is often tightly specified, and even made to be  
like a constant current source for some tubes.
- The 3-9 mc "ARC-5" receiver I had as a novice in the late 50's still  
has its original tubes installed.  I would expect them to work fine  
now (even though I have not run it for a long time.)
- The EIMAC publication "Care and Feeding of Power Grid Tubes" has a  
section on filament management for transmitting tubes.
- To test radio performance with respect to filament voltage, I  
suggest testing with a variac  on the filament supply to detect a  
noticeable reduction in audio output, AVC off.  The above article  
indicates that when emission limited performance is reached,  
presumably due to long service or low filament voltage, the noise of  
the tube increases.  Maybe a measure of the noise of a receiver would  
be a good test when investigating low filament voltages.
- Reference to the RCA Handbook 3 sections on tube life- and  
performance- testing will shed light on filament management for  
receiving tubes.
- I think that receiving tube failure/end-of-life is more often due to  
reduced cathode emission than to filament failure.
- I've noticed that European tubes of the dual triode sort (e.g. 12AU7  
and 12AX7) start up with a brief brightness seen from the exposed  
filament wire whereas American made ones don't.
- For those who want to try out the WE=416A or B, www.tubeworld.com  
has them for $23 to $25 each.  Not bad compared to the $125 or so they  
want for the 417A.

Roy


Roy Morgan
k1lky at earthlink.net
K1LKY Since 1958 - Keep 'em Glowing!