[Hammarlund] Tube shields and heat

[Richard Knoppow]

   I pointed out earlier that since the tube has a hard 
vacuum in it the transference of heat from the elements to 
the envelope is almost completely by radiation. Think of a 
thermos bottle here, the main insulation is the vacuum 
inside the double envelope. Some heat probably radiates back 
to the plate but, in general, cooling the envelope will not 
have much effect on the elements.
   Once in a while you see a picture of a transmitting tube 
with the glass sucked in, Eimac used to run ads like that. 
That is from local heating of the glass due to electron 
bombardment, not heat from a glowing plate. Its possible to 
get the glass hot enough to cause problems with the seals 
where electrodes come through the glass. The blowers used on 
old style glass transmitting triodes were for cooling the 
seals rather than the tube's internals. External anode tubes 
are, of course, different because the heat can be removed by 
conduction.


--
Richard Knoppow
Los Angeles
WB6KBL
dickburk at ix.netcom.com
----- Original Message ----- 
From: "Bob Camp" <kb8tq at n1k.org>
To: "Hammarlund QTH LIST" <hammarlund at mailman.qth.net>
Sent: Saturday, October 04, 2014 5:36 PM
Subject: Re: [Hammarlund] Tube shields and heat


Hi

Chemical rate law phenomenon are one of many things that 
cause degradation in a device. The rate law is true once you 
reach an activation energy level. For many of the materials 
in a tube that occurs in the > 1,000 C range. Yes a lot 
depends on exactly which reaction you are tracking on each 
material.

Cooling the outside of the tube does little for things like 
the filament. Its going to be at glow temperature regardless 
of what you do on the outer envelope. The grid to cathode 
spacing for a tube to work is such that the control grid 
will pretty much follow the cathode, regardless of what you 
do to the plate. The only thing you may be cooling with a 
black shield is the plate and to a lesser degree the 
suppressor grid (if its a pentode).  I have never seen a 
tube fail for glass envelope degradation due to temperature. 
Dropping them does not count in this case. Doubling the life 
of the glass envelope, when it never falls regardless does 
little to improve the tube. Glass to metal seal temperates 
are a different issue, they are little impacted by a tube 
shield.

Since you depend on heat to activate the getter material, 
and its on the tube shell, cooling the shell may actually 
degrade the reliability if its taken to far.



Why do most tubes fail?

1) The filament / cathode system runs out of coating / 
electron radiation capability. It stops glowing, at least in 
terms of electrons.

2) The envelope fills with gas, but that does not poison the 
filament. This happens when a glass to metal seal fails.

3) A weld breaks and the tube just goes dead. Normally this 
is due to cycling the tube. Welds are a high activation 
energy item.

4) A grid slumps and it shorts to another element in the 
tube

At least of the few 10s of thousands of tubes Ive seen, 
number one wins out by far. That one is not impacted by 
cooling at all. Numbers 2 and 3 depend a lot on what sort of 
tube it is. They happen more often than number 4 by far. I 
have yet to see a receiving tube with a hole burned in the 
plate. Maybe it happens, if it does, its rare (so its not on 
the list).

Cooling the tube isnt going to impact 1 or 2 (the big ones) 
at all. It may impact 3 depending on the location of the 
weld. It certainly would impact 4 on the suppressor (but not 
the cathode or control grid). Holes burning in the plate are 
the big one that cooling would help. Since you arent 
addressing the main wear out mechanisms, cooling isnt worth 
it.

Bob