Tube Failure

[jwatts]

My question is. What can go wrong in a tube that will cause it to not function or change its normal operating parameters? I suspect the filament can burn up. I thought maybe that some type of carbon buildup could occur on the plates. Tubes are under a vacuum so nothing is there to buildup. Can any body give any input on this thought?

[Sherman]

Originally posted by Jason Watts
... What can go wrong in a tube that will cause it to not function or change its normal operating parameters?...

First, like a lightbulb the vacuum can fail. This can result in a rather dramatic failure.

Second, on a more "theoretical" note, cathode stripping is possible. When high voltage is applied before the filaments are heated sufficiently emissive material can be "blown off" the cathode. (Keep the volume control fully attenuated until operating temperature is reached on SS rectified amps.)

Third you can have mechanical changes due to impact which could cause a change in the cathode/anode distance and which could impact operating parameters.

Fourth, as you surmised, filaments do not last forever even if the vacuum remains strong so you may get improper or uneven heating.

Fifth, oxidized pins (or a related problem but not one with the tube itself- oxidized socket contacts) can affect voltage and current at all important points- cathode, grid, plate.

I'm sure there are lots more but those are off the top of my head.

[GRollins]

Besides breakage, you mean?
There's no carbon in a tube, so you can set your mind at ease about that one.
Let's take it element by element:
Filament--the filament is actually pretty durable as long as you don't run it at too high a voltage/current. Over time, it will emit fewer electrons, but the degradation is slow.
Cathode--this surrounds the filament and in some tubes is absent altogether. Most of the common audio tubes have a cathode that is separate from the cathode, however. Like the filament, it's ability to emit electrons will drop over time. Cathodes have a maximum current rating. If you exceed that, you can damage the tube.
Grid--Under normal conditions, the grid will last nearly forever. However, if the tube arcs internally (e.g. a power pentode), a portion of the grid will be destroyed. As you might imagine, arcing isn't too good for the rest of the tube, either.
Plate (aka anode)--The plate has a wattage rating; how much heat it can safely dissipate. If you exceed this, the plate can get to the point where it is literally glowing cherry red. Not good. The heat deteriorates the plate and releases gasses which were chemically bound up in the metal. This causes problems with the...
Vacuum--the vacuum will fall very, very slowly over time due to an imperfect seal around the tube pins or outgassing by the metal tube elements (and a little bit from the glass). In a well made tube, the leakage is so slow that you can disregard it.
Getter--the silvery or dark splotch inside the tube is a mixture of metals that react quickly with gases...but not all gasses. Oxygen, for instance, is removed very quickly. Hydrogen is not as easily dealt with. The getter deteriorates with the vacuum. The more gas comes in, the more quickly the getter deteriorates.
Glass envelope--the glass, with the exception of the seal at the pins, is pretty much good to go unless you break it. It does contribute a small amount to the outgassing problem.
Pins--corrosion or breakage.
Tube base--breakage.
As a laundry list, it looks pretty bothersome, but as long as you treat the tube well, you'll be all right.

Grey

[richwalters]

[What can go wrong in a tube that will cause it to not function or change its normal operating parameters? ]

Hi there......bad circuit design....overloading g2...is quite common; many tubes (even power ones) can only take a few mA/Watts. An o/p tube may still work with a damaged g2 but the beam will skew and create high thd.
Duff indirect heater is rare unless physically dropped to create a heat/cath short.

Directly heated cathodes are more suspectible to damage with voltage variations. Tendancy for 5U4G rectifiers is gradual peeling off of the emission coating from the heater elem over time.
Next time take a closer look at your toaster elements.....same thing going on.
I've repaired juke box amps using EL84's and 6L6's in high vibration proximity to the LS...... and still worked for decades under these conditions.
One takes what on gets. Check the circuits currents.....Replacement tube is the only cure.


rich

[SY]

Tubes are Murphy encapsulated.

Cathodes strip and poison. Envelopes get gassy. Spot welds separate. Grids sag and get poisoned. Micas flake and break. And on and on.

[jwatts]

What I was really trying to find is. If a tube is going bad how do I test it? I thought possibly that the first thing you would look for is check and see if there is continuity across the filament. Soemthing tells me it is a bit more complicated than that.

[SY]

Aha!

The best way is to actually test it. The first thing is indeed to note whether or not it's glowing. You then want to check transconductance, gain, grid leakage, noise, microphonics, and distortion. For some uses, you may need to check more things, but if you have this stuff in hand, you can deal with 99.9% of the uses in audio.

None of this stuff is difficult or expensive to measure, but you will need to throw some test jigs together.

[jwatts]

I have all the basics multimeter, oscilloscopes, spectrum analyzer, etc. Can you please direct me to a location on the Internet or a book describing on perform the necessary measurements and construction of the necessary test equipment.

[SY]

I'd start with the basic books that anyone serious about tubes should own- RCA tube manual, Morgan Jones "Valve Amplifiers" and "Building Valve Amplifiers", and Radiotron Designer's Handbook. For an example of the sort of measurements you can do with a decent (but not extravagant) sound card, search for my post on 5692 measurements, "SY gets jiggy."

[jwatts]

Assume a tube begins to slowly fail and the properties of the tube slowly begin to change. How does this affect the attached circuitry? I mean can the failure of a tube damage surrounding components directly connected to the tube.