Why do power tubes degrade?

[spacelabstudio]

Here's my Saturday night "I wonder why?" question.  Conventional wisdom seems to be that power tubes "wear out"--their performance degrades over time.  But why?  What is it, physically that is wearing out, or degrading? 

[proaudioguy]

Lots of heat and lack of pure vacuum?  Vibration?

[Willabe]

Short answer? From what I understand about it.     

Just like a battery, a tubes cathode only has so many electrons to give up and then no more.    

There are other things that will wear out/kill a tube and/or shorten it's life, like heat or over stressing it by running it to hard. Kinda all the same thing in the end, just makes it happen faster.

In a perfect world you would have a perfect vacuum and perfect materials but that's not gonna happen in this life.


               Brad            

[FranciscoPerez]

My personal experience is that in practice power tubes degrade from mechanical reasons (vibrations etc) the most, in case they are not pushed too hard.

[spacelabstudio]

The question, really, is what changes physically inside of the tube that alters its performance?  The external factors that can lead to degradation seem to be widely known, but why those factors have an impact, through what physical mechanism, seems to be harder to find.

[sluckey]

The cathode is coated with a material that will easily release electrons at a moderate temperature, ie when heated by a filament placed nearby. Over a period of time, sometimes decades, that coating will literally be baked away. As more and more of this coating is destroyed, flaked away, powderized, vaporized, etc., the cathode emission (release of electrons) decreases. The tube is said to become weak.

Other physical problems can occur too. Grids can sag over time, changing tube characteristics or even shorting to other elements. Elements can become loose or detached from their mounting and short to other elements. Tubes can become gassy or lose their vacuum.

[HotBluePlates]

Download and read Getting the Most Out of Vacuum Tubes by Robert Tomer, 1960.

Over time, cathode emission drops somewhat, but that takes longer than you'd realize. Tube characteristics can degrade due to contamination of the grids from the cathode coating during use (or more often, abuse), gas can be liberated from the metal elements, the getter can be depleted and fail to perform its function, etc.

But the folks who tell you power tubes wear out in 6 months are either selling you tubes, or repeating the statements of tube vendors. Ditto about phase inverters needing to be replaced with the output tubes.

You may hear changes in the tone of your output tubes, but that's most likely in amps with no negative feedback around the output section. The true test is to feed a test signal to the output stage and measure the actual power output, changing the tubes only when the output has dropped below some lower limit.

Reality is you can probably keep using a given set of tubes until the filament burns out or you don't get anything from the speaker. Unless something else kills them before that. I have a couple sets of 6V6's that I've used in various amps for more than a decade, and they still sound good and work well enough for me. I haven't measured their actual power output, but I can irritate the neighbors if I need to...

[jjasilli]

Ditto re the cathode.  But I think the main problem is that the heaters burnout, just like the filament in a lightbulb.  My 6550's usually fail because the getters turn white.  The cathode is supposed to give off electrons, but unfortunately it also emits ions -- gas molecules.  This degrades the vacuum in the tube.  The getter(s) absorb the gasses but only for a limited time.  They start off silver, gradually turning white; and leave a white dusty residue inside the tube.  The tube should be replaced before it fails.

Plates & screens draw current and expend watts, giving off heat.  They can burnout too.  Or expand & short-out against some other internal component, causing tube failure.

EDIT looks like hotbkue beat me to the punch!

[sluckey]

Download and read Getting the Most Out of Vacuum Tubes by Robert Tomer, 1960.

Now that's a complete answer!

[spacelabstudio]

That's what I'm talking about gentlemen.  Thanks for some interesting info to chew on!

[HotBluePlates]

My 6550's usually fail because the getters turn white. ...  They start off silver, gradually turning white; and leave a white dusty residue inside the tube.

This statement scares me a little.

I've only seen white getters when the tube has lost all vacuum, such as a broken envelope, or a pin bent so badly it cracks the glass seal. Normally, you see the getters fade at the edges and eventually disappear as they combine with the gas molecules in the tube.

Download and read Getting the Most Out of Vacuum Tubes by Robert Tomer, 1960.

Now that's a complete answer!

People under a certain age (me included) might not realize that CBS was a manufacturer of vacuum tubes, until you get some CBS-labeled tubes. So it's worth noting that Tomer wrote the book from the standpoint of a person intimately familiar with vacuum tube manufacturing.

On the topic of gas, he points out that an absolute vacuum is/was impossible to create in tubes. But the volume of gas in a tube was so minute as to be largely a non-issue, except for the mechanisms by which gas is liberated from tube elements.

In any event, I've got 1,000+ tubes (lost count a number of years ago). I've had only 1 short-circuit in almost 20 years of playing with them (heater-to-plate short from pin 2 to pin 3 of a 6L6GC). I've only had a couple with white getters, generally due to breakage in shipping or from dropping a box holding a bunch of loose tubes.

Outside that, I've never had a tube with the heater open up, and I've never had a completely dead tube that couldn't be made to provide some amount of function in circuit. Gain, Gm, or power output may be low compared to a new tube, but they all still do something. So it really boils down to how you evaluate their performance, and at what point you decide to pitch them.

[jjasilli]

Heaters:  I've had a number of inoperative tubes in which the filaments did not glow.

Getters:  see http://www.audiotubes.com/nos.htm & http://thetubestore.com/blueglow.html 

I can't find a certain article I've read on buying used or supposedly NOS tubes.  One buyer-beware problem is that the tube may look & test OK now; but if the tube were in service for a long time, the getter may be secretly about to enter its death spiral and soon show signs of decay.

[sluckey]

I replaced a lot of tubes that had open filaments back in the days when I worked in a radio and tv shop. Open filaments was a common failure in series strings of filaments. However, I left the tv shop in '69 and entered into the radar field. I cannot recall ever replacing a receiving tube with open filaments during my 40 year career. High power transmitter tubes was another story though.

[PRR]

White getter is almost always air, cracked seal.

By design, the getter will hold about all the smog which could be released from the internal parts.

I have not done extended abuse on recent-made tubes. Golden-age tubes had pretty good seals, as long as you respected the maximum temperature (by respecting dissipation rating), but they WOULD fail if the tube got over-hot; that's a key limitation of the way receiving tubes are made. (Transmitter tubes use different glass, seals, leads, but cost a LOT more to make.)

If you are whiting-out 6550s regularly, either you are abusing them or the modern makers just don't make seals like they used to.

I did once cook a Sovtek 6550 at full 42 Watts for a couple days, no ill effect. That was steady heat; constant hot/HOT!/hot cycling as in an over-loaded music amp might be worse.

Series-TV sets did burn heaters; otherwise my experience is like HBP, heater failure is nearly unknown.

[jjasilli]

If you are whiting-out 6550s regularly, either you are abusing them or the modern makers just don't make seals like they used to.

I have been buying used 6550's -- GE , RCA, Sylvania, etc.  in my 3 home theater system PA amps -- vintage Stromberg Carlson APH 1100's.  They run, say, 3 hours per day; about 20 hours per week; 1000 hours per year as a guesstimate.  600V Plate plate; 300 screen; 54 mA.  I have not kept good records but I'd say the tubes last abut 6 mo to 2 yrs. (New JJ Kt-88's / 6550's have had a worse failure rate!)

[alerich]

Download and read Getting the Most Out of Vacuum Tubes by Robert Tomer, 1960.

What a great little website. I actually have a hardcover copy of one of those books:

"Inside the Vacuum Tube" John F. Rider, 1945, 407 pages

but mine is a seventh printing from 1952. Thanks for the link. Much appreciated.

[Ritchie200]

Don't forget amps that are running high screen voltages that are pushed hard by the preamp.  They are not long for the world when beat up. 

jjasilli, I am not surprised that the JJ's are not lasting in that application.  Try some ~C~ KT88's.  THey hold true to the old KT88 architecture.  I can still run those all day long and they sound good - and I've got them running pretty hot.  The best replacement I've found for the old Genelex.  I've had numerous and frequent failures with other brands.  Much of the new stuff (in KT88's and 6550) pale in comparison (IF they do hold up) to even well used old stock in high voltage applications.  That sounds like a GREAT system!

Jim

[jjasilli]

Ritchie200:  I'll PM you to avoid further thread hi-jacking.