heater voltage variation for special tubes

[nazaroo]

I have seen the RCA chart that shows drastic reduction in tube life and performance when you stray from recommended voltages by more than 5%.

Although you can go further decreasing the voltage than you can increasing it, obviously there simply isn't much headroom at all.

This doesn't seem related at all to cathode-stripping per se,
and nor does it take into account smart slow-start circuits that put less stress on the heater filaments.

I would think that you actually COULD lower the voltage a bit more if you have a slow-start warmup circuit in place. It would be nice if anyone did a study however, taking into account that parameter.

I am thinking you could drastically increase the life of a tube with the slow-start, then lower the distortion by using less voltage at the same time, and you'd actually be AHEAD in terms of tube-life, trading some off for better control of emission, while taking advantage of gains from the slow-start.

Secondly, I don't think these charts by RCA can apply to direct-heated cathodes for big tubes. I'm thinking either thoriated tungsten or large carbon cathodes might not be affected by voltage fluctuations or even current fluctuations very much, at least within a much wider spread than smaller modern tubes.

For instance, in the Taylor manuals, they mention (boast) about the power handling of their large cathode powertubes. Its as if the voltage was almost irrelevant for these big tubes.

The reason I'm interested is I want to use some 813s, and 814s, (they have different cathodes btw), and I am wondering if I can get longer life from them by really lowering the heater currents.
The tube-makers are suggesting a VERY wide range of Plate voltages and even Screen voltages for some of these power tubes. If this is the case, and if this doesn't really affect tube-life, (other than conservation of cathode emission capacity), then it seems that with such tubes we can have quite a lot of headroom to get extraordinary tube-life.

In these days when these large rare NOS tubes are no longer being made, tube-life may be one of the primary considerations in modern circuit design using these tubes.

What say ye?

 

[DF96]

You seem to be conflating the wide range of anode voltages and currents with the narrow range of heater currents. For most valves, heater currents need to be correct. For long life you should reduce anode voltage and current, and keep the envelope cool.

 

[nazaroo]

You seem to be conflating the wide range of anode voltages and currents with the narrow range of heater currents.

No, what I am asking is, is there a relation between the range of anode voltage and the robustness of the heater element in these older, e.g., directly heated cathode style tubes?

I think there must be a perhaps unnoticed correlation:
That is, tubes with large carbon heaters for instance,
and which subsequently have larger ranges of anode voltage (in part due to size and power),
will also have larger ranges of heater voltage/current operation.

For most valves, heater currents need to be correct. For long life you should reduce anode voltage and current, and keep the envelope cool.

But the question I am raising isn't about "most valves", but about a specific group of valves, those older, larger powertubes with carbon heaters and/or thoriated tungsten.

You've just made two contradictory statements here,
because you yourself have conflated what is true for modern valves (e.g., RCA indirectly heated signal tubes, rec. tubes and modern power tetrodes), with what is patently not true for older valves.

You can always reduce anode voltages and screen currents to increase valve life.

What I'm pursuing here is special properties of older valves.

 

[SpreadSpectrum]

I remember reading somewhere that thoriated tungsten filaments are much more resistant to ion bombardment than oxide coated filaments. For instance, compare max plate voltage ratings for an 801A and a 45. I'm assuming electrode spacing isn't much different on these two, but I don't have them to compare. I've just noticed that max DC plate voltage goes up quite a bit for TT filaments even when the tubes being compared are the same size.

Oxide coated cathode tubes may need to be in the sweet spot more to protect from ion bombardment. Just speculation.

 

[HollowState]

I remember reading somewhere that thoriated tungsten filaments are much more resistant to ion bombardment than oxide coated filaments.

http://www.diyaudio.com/forums/tubes-valves/209294-valves-tungsten-filament.html

With thoriated tungsten filaments, it is important to maintain filament operating temperature by rated voltage and not by trying to set current. The filaments will draw what current they need. And filament current in these types of tubes is a telltale sign of their condition according to Eimac. As a TT tube ages, carbon is drawn to the filament surface. This lowers the resistance and increases current draw. At around 5% above bogey, the filament is considered to be at end of life. (or very near it)

I'm not sure how closely this relates to regular non-thoriated filaments, but I never current regulate directly heated cathodes for this reason, even though it seems to be in fashion with some people.

 

[nazaroo]

http://www.diyaudio.com/forums/tubes-valves/209294-valves-tungsten-filament.html

With thoriated tungsten filaments, it is important to maintain filament operating temperature by rated voltage and not by trying to set current. The filaments will draw what current they need. And filament current in these types of tubes is a telltale sign of their condition according to Eimac. As a TT tube ages, carbon is drawn to the filament surface. This lowers the resistance and increases current draw. At around 5% above bogey, the filament is considered to be at end of life. (or very near it)

I'm not sure how closely this relates to regular non-thoriated filaments, but I never current regulate directly heated cathodes for this reason, even though it seems to be in fashion with some people.

Thanks for this headsup and explanation.

I don't see any reason here that prevents lowering the voltage however,
but maybe I'm missing something.

In any case, one should be able to lower the voltage more
than one can with an oxide-coated or modern tube...

Just wanting to nail this down for purposes of circuit design.

 

[trobbins]

So is it that the increased current draw with aging would lead to a 'walk-away' heater failure, given time?

If the heater supply to the particular tube was actually from a constant 'power' arrangement then the heater could continue to do its job, without regard for change in DC characteristic - or was there another degradation of the heater performance that dictated replacement?

 

[HollowState]

I don't see any reason here that prevents lowering the voltage however, but maybe I'm missing something.

I'm of the opinion that, with rare exceptions, tube filaments should be operated at their design center rated voltage. Wandering off that value can put the tube's useful life at risk. Back when tubes were manufactured by companies that knew what they were doing (sans todays Chicom junk), I choose not to second guess their wisdom. Chemical actions in the filaments (cathodes) were designed to operate at specific temperatures that were determined by the applied voltage which was the primary criteria. The current consumed being secondary.

I have seen instruction sheets for new special oxide cathode transmitting tubes by RCA and Burle that say to reduce filament power until output just starts to drop, then go back up a little bit. And thereafter as power drops, increase the filament to compensate. But this is for a special class of tubes and should not be applied to the tubes we use.

I've also seen (in test equipment) some small tube types (12AX7's) operated at reduced filament. This was done to lower noise levels in low level circuits, not to increase longevity. It was a design trade-off. I'm not sure it the life was extended or not from running cooler. Logic makes one think it would, but the cathode coating may ultimately suffer poisoning from running under temperature.

 

[HollowState]

So is it that the increased current draw with aging would lead to a 'walk-away' heater failure, given time?

I'm not exactly sure what you mean by this. As a TT filament ages, it slowly looses emission. And as surface carbon increases, it becomes more brittle. Then at some point emission becomes low enough that replacement is necessary.

Now, there are many small broadcasters that, instead of replacing the tube, they'll crank up the filament voltage. This will continue until every last electron is sucked from the filament. And the RF driving power can be increased to force more output. But since the filament power required for big transmitting tubes is large, true regulation is not used. At lease I've never seen it. However, the tube manufactures always state that the filament voltage should be tightly controlled. Usually this is done with a quality voltmeter and a variac. Either manually or automatically by motor drive.