[Amps] General question on filament life: comments

[TexasRF at aol.com]

Wonder if any study has been given to the life of these lamps when the turn  
on current is limited. In other words, would the "about ready to fail" time be 
 increased by providing some turn on current limiting protection? Maybe the 
turn  on cycles wears the lamp out more than the actual on hours?
 
No one would ever do that to help the life of a 50 cent light bulb but how  
about a $1000 transmitting tube?
 
I run a TH347 tetrode on 1296MHz that qualifies (begs) for some filament  
protection. The tube manufacturer warns in the data sheet to limit the filament  
current to something like 2X for the first a.c. cycle. Taking this a step  
further I decided to limit the current to less than 1X at turn on. This was  
accomplished by connecting a couple of 100 ohm surge limiters in series with the  
filament transformer primary. With 200 ohms in series, a maximum of .6 amps  
flows in the primary and 12 amps in the secondary. The tube is spec'ed for 34  
amps filament current so this is very conservative current limiting. About  
three seconds after turn on, the filament volt mater begins to rise and reaches  
about 5.8 VAC in maybe five or six more seconds. At this point, a Dayton 
Solid  State Timer connects AC to a small DPST relay. One set of relay contacts 
shorts  the surge limiters, allowing the full 6 VAC to the tube filament and the 
other  set closes a control loop for hv and ptt circuits.
 
Shorting the surge limiters allows them to cool to cabinet  temperature, 
making them return to 200 ohms and ready for the next turn on  cycle. Also, I 
assume that the lifetime of surge limiters is extended by  not allowing them sit 
and cook for hours on end. 
 
This is so simple and inexpensive that I intend to add the circuit to some  
other 8877 VHF and HF amplifiers that normally don't use filament surge  
protection.
 
If anyone is aware of a downside in using such a circuit I would appreciate  
hearing from you.
 
73,
Gerald K5GW
 
 
 
 
In a message dated 2/11/2007 3:47:59 P.M. Central Standard Time,  
g0fvt at hotmail.com writes:

Hi,
not much experience of this, however I once worked for a lamp  manufacturer, 
there was a tendency for lamps to fail at switch on ( I  believe we have all 
seen this).
The cold resistance of a typical  tungsten filament is roughly 14 times lower 
than the running resistance so  if the supply impedance is low the inrush 
current is in the region of 14  times higher than the running current. I 
think the failure mechanism is  related to the uneven rate of temperature 
rise whereby some parts of the  filament briefly run much hotter than normal 
as some other parts are still  warming up. I believe I have seen this in 
tubes/valves whereby sometimes  the filaments go quite bright on switch on 
and subsequently appear to go  dimmer.

With the life testing of lamps it was acknowledged that lamps  that failed on 
switch on had little life left anyway, I believe the life  testing schedule 
called for one interruption to the power source for 30  mins each day to 
simulate a reasonably realistic  situation.