If a new electron tube has a build in mechanical defect, it will probably occur within the first few hundred hours of use. To give an example: micro cracks in the glass, will not give a problem during decades of storage, but may develop a vacuum defect after some 500 hours of first use. When a tube will stay good for this time, you can expect such a defect will not develop at all. For many other defects, the effect is similar. The situation with guarantee on vacuum tubes can be disappointing when vendors give only 3 months guarantee. This is often to short, to reach 500 hours. For instance, when a tube is used for 12 hours per week, it would take 42 weeks to reach 500 hours, and the 3 months guarantee would be already over. To make sure, all defects are covered by the guarantee, we have a standard guarantee time of 12 months or 2000 hours (whatever comes first) , and it can be extended to 5 years by internet registration. Please note this: Apart from visible mechanical failures, the major defect mechanism with older tubes, is a vacuum defect. For this reason, at EML we include vacuum defects in the 5 years enhanced guarantee. (See below text for the guarantee conditions). Upon registration, the buyer will receive a personalized response, with the tubes series numbers, and all relevant details. The email address will only be used to confirm your registration, and we delete all data if no longer needed.
With all tubes, regardless the brand, high plate dissipation, or high current, will result in lower lifetime. The safe operation area is below 75...80% of maximum plate dissipation, and "long life" application is around 60% of maximum. In some special cases, a tube can be used at maximum power, and lifetime is still maximized as well. With such tubes it can be seen in the data sheet, when the typical dissipation and maximum dissipation is the same number. For instance the RCA 2A3 is specified like this, but Western Electric 300B is not. So you can use RCA 2A3 (and EML too) at full maximum dissipation, and lifetime still will be highest. Western Electric 300B (and EML too) will benefit from staying below maximum plate dissipation, same as most other tubes.
The best lifetime will occur at the specified heater voltage, a safe electrical circuit, no crude start up, and no oversized capacitors for rectifier tubes. Also coupling capacitors can shorten lifetime if dimensioned for a few Hz only, which can create positive grid current during start up.
The plate current is factory tested after initial burn-in. This level is called 100% when the tube is new. This is an individual value for each tube, and the required grid voltage for this is written on the tube box. If you wan to judge the condition of an older tube, you should test it at just that same Grid voltage, as we measured when the tube was new. It will stay at or close to that level during customer burn in. In the beginning the plate current can change (up or down) very slowly, or stabilize for most of the time. After it stabilizes, the customer burn in is finished, and normal wear out will start to begin. You will observe the plate current going down very slow, each few hundred hours. So if the plate current (referring to the original test value) has decreased, this is a normal sign of use, and no defect. This will continue during the entire lifetime. Amplifiers with Auto bias, will hold the plate current fairly good. Amplifiers with adjustable bias, will need regularly adjustment. That's what the adjustment is for, because these have no Auto bias. When the end of lifetime is getting nearer, the plate current will start to go down quicker than before. Generally, you can say when the tubes sound good, that's what they are. Only you can not judge the remaining lifetime in the tube by that. On a test bench, the measured values of fairly used tubes can be at 70% plate current. There is a general conception that such tubes have still enough lifetime left in, to make a well designed amplifier function normally. This value of 70% is found on most tube testers as the level where the "good" reading begins.
End of Lifetime
This becomes audible, when the tube starts to distort at loud music, and was not doing so before. When you can measure the plate current, the area for a "good" tube is above 70%. Below this is not always the end of lifetime, or a bad tube. Tubes with values of 40% have been reported to work normal, depending on the amplifier. Below 40% problems will probably occur. So you can feel safe when buying used tubes at or above 70%. Not below that.
What causes long tube life?
Provided, the manufacturer supplied a first class product, the user can contribute by respecting the condition of the cathodes (the heaters).
During normal use, the cathode must be so hot that electron emission takes place at a high level. Unfortunately this temperature is so high, that a small fraction of the Barium layer will evaporate by this. This gives small openings in the layer, and at these openings, the Barium Oxide layer underneath gets exposed to electrolysis. The result of the electrolysis is Barium Oxide (white powder) gets reduced to metallic Barium, and Oxygen gas. The Oxygen finds it's way to the getters, and the Metallic Barium closes the hole. This process is like a self repair. However after some time , the Barium Oxide layer gets depleted.
A carefull balance must be found.
So the cathode must be hot enough to give the emission, and also more heat gives more regeneration (calles electrolysis before). However the heat will also cause evaporation. So one way or another, wear out is unavoidable. So the regeneration must be as little as necesairy, but as high as needed.
Once this process is understood, it is clear we want two things, which seem to exclude f each other:
Keep Barium evaporation at a minimum level. (Reduce heat).
Create a self-repair process. (Increase heat).
In other words, over heating may seem to work nice initially but it will wear out the tube much faster. Under heating will greatly reduce wear out, but there is not enough regeneration. It comes down to using the correct heater voltage with as little tolerance as pissioble.
Any questions, like what heater voltage tolerance is allowed, we answer always like this: Zero tolerance is the best!
Over heating will must faster evaporate the Barium metal. Furthermore, over heating will speed up the migration of Barium from within the cathode. So from electrical testing, it seems as if over heating doesn't matter much. The tube seems to take it, and tests normally good afterwards. Yet this impairs lifetime very much. Overheating will empty now the Barium depot very fast, and when it is used up, migration will stop, the Barium layer will evaporate finally, and this time without repair. The tube life is over suddenly and unexpected.
Under heating. If the cathode is not warm enough, like when heating a 5V tube with 4.5V only, this is very good to reduce the Barium evaporation, and initially it seems like a good idea, since the tube works still normally, even below 4.5V. So users do not understand why exactly 5.0V is important. However the balance is disturbed, the migration in Barium to the surface is now a LOT lower. After weeks of use, the Barium layer gets too thin, and holes begin to appear. Now, soon emission will go down, the tube will sound distorted, and test "defective" on a tube tester. If corrected timely, this damage will repair by normal use, at exactly the right heater voltage again. However if the loss of emission was only corrected by adjusting the bias, this will cathode damage becomes too large, and the tube damages permanently.
Balance. This the key and the secret of long tube life. So use just the right heater voltage, do not overheat or under heat the cathode. Do not overheat the anode, as this will effectively overheat the cathode as well by back radiation to the cathode. Do no draw excessive current peaks from rectifier diodes by using too large capacitors. Do not overdrive Push Pull amplifiers just for "fun" to hear what this sounds like. All of these things will blow holes in the Barium layer, which need continuous repair. This uses up the depot faster.
Warm up. Make sure you have no circuits that draw peak current while the heater is not fully warm. Very infamous are rectifier circuits which by default have the full AC voltage from the transformer, but are loaded by an empty capacitor at switch on. So oversized capacitors do extra harm here.