Amps
[Top] [All Lists]

[Amps] Dimished Tube Life versus emission

To: amps@contesting.com
Subject: [Amps] Dimished Tube Life versus emission
From: "John Lyles" <jtml@losalamos.com>
Reply-to: jtml@vla.com
Date: Fri, 25 Sep 2009 10:32:29 -0600
List-post: <amps@contesting.com">mailto:amps@contesting.com>
This reply is based on my own experience, plus talking to various tube 
manufacturers over the years, dealing with larger power tubes. Emission 
lifetime, of course, depends a lot on how the tube was set up, was the filament 
burning at the rated voltage or turned back to preserve life? Also on how much 
gas evolved in the tube during operation. This could be either from the 
original manufacturing, or from overloading the tube elements at high average 
power. Stray ions can damage the emitter and reduce its life. 

Lets stick with one type of emitter for this answer, thoriated tungsten (TT). 
This means this discussion doesn't necessarily apply to 8877/3CX1500A7, 
4CX250A, and those sorts of tubes having oxide cathodes. Someone else can reply 
about those:
With TT, the filament is carburized with a layer of carbide during processing, 
to greatly enhance emission from a single molecule-thick thoria layer in the 
filament.  This coating, plus the thorium percentage in the tungsten, are 
factors that give long lifetime. If the tube becomes decarbed, where the 
carburization is depleted, then it will promptly run out of this level of 
emission that is required to support the desired cathode current. Actually the 
emission would just fall off a tremendous amount. The only way to recover it 
would be to raise the temperature much hotter, which will physically destroy 
the filament. So it is a loosing game at this point. Usually it is not the 
depletion of thorium that causes the emission shortfall. 

When emission starts to droop, the usual solution to squeeze a little bit 
longer life out of it is to raise the fil. voltage. This, indeed will improve 
things for a short time. Then it slumps off again, and the voltage is again 
raised, and so forth. Usually its a matter of weeks to months for a large tube 
operating 24/7 before it cannot be increased anymore, the filament setting is 
too high, power supply ran out of range, or the filament breaks. This also 
becomes  a nuisance, having to continually tweak the voltage. If you just 
turned it up, say 10%, you would see a burst of emission, but then the lifetime 
would be reduced greatly in terms of time duration before failure. Don't do 
that. 

Now this brings up a question of "what does 75% output mean". In a particular 
circuit, the designer could choose to use an oversize tube, that has higher 
emission and cathode current than is needed for the circuit. This would be a 
costly practice, but is sometimes used in pulsed power systems (such as where I 
am at work) where a big tube with lots of plate dissipation is used, not for 
the plate dissipation, but for the higher cathode current (and plate current as 
noted on characteristic curves from the datasheet) during the pulses. In this 
case, extreme long life can be had by reducing the filament voltage a bit,  on 
the order of years. A used tube pulled from this application may have a lot of 
life left in a lower power system. On the other hand, if a particular circuit 
designer chose a tube close to its ratings, for peak current and for 
dissipation, the filament may need to be near its nominal rating, which means 
the lifetime will usually be somewhere on the average like 10
 ,000
hours or less. In this case, when the tube performance degrades to 75% power, 
then there isn't much left to go on. The tube will rapidly decline to the point 
where 50% or less is all it may do. So this question of 75% output, really 
reflects on how the original circuit was running this tube. A pull from an AM 
broadcast transmitter modulator may be really spent, as they run them until 
distortion creeps up or cannot reach 100% AM anymore. 

There won't usually be a short that ends life for a decently designed tube and 
circuit. It will typically be fall off of emission until the peak plate current 
can not longer be reached at the peaks of the RF or audio waveforms. If a TT 
tube is run for many years, however, the TT filament does become aged and 
brittle and is subject to breaking if the tube is handled roughly while the 
filament is cold. Another thing that may indeed cause a short is to try and 
raise the screen voltage for a tetrode, or drive a grid harder for a triode, to 
try and compensate for reduced output due to emission decline. This can cause 
an overheated element which may indeed short to another element and kill the 
tube. 

This reply is not straightforward, but neither is end-of-life for a tube. A lot 
depends on the circuit, how the tube was handled, how close to rated current, 
filament voltage regulation, etc. When a tube starts to fall off in power, and 
it is closed rated for the circuit design, then it will continue to fall off 
for some time. If this is not the case, then you should be able to bump up the 
filament voltage to get more output, at least for a while. There is no single 
answer that covers all tubes with your question, a lot depending on the circuit 
and the conservatism of the engineer who designed it and selected a particular 
tube. You can continue to use it to zero output, but the decline will be fast 
at this point, as it will inevitably become de-carburized. Good luck with that. 

73
john 
K5PRO


> Message: 7
> Date: Fri, 25 Sep 2009 11:05:38 +0000 (UTC)
> From: Paul Decker <kg7hf@comcast.net>
> Subject: [Amps] Dimished Tube Life
...
> When is a tube considered at it's end of life??? Of course a tube like I 
> posted about last week, where there is a grid - cathode short, is (at least 
> right now) at its end of life.?? But if over time the tube is simply having 
> reduced output or "going soft", at what point is it considered "no good"; at 
> 95%, 80% 75% output????? 
>
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>