As there is a quite a bit discussion on this thread about the process of vacuum tube “burn in”, I thought I should bring up the (hopefully obvious) fact that vacuum tubes are a mature technology and is, at this point in time, reasonably well understood from an engineering standpoint.
Not only is tube burn in real, back in when this technology dominated much of the electronic industry, companies wouldn’t ship vacuum tubes to the customer until they had been thoroughly burned in at the factory (they had machines that could burn in hundred or, in some cases, thousands of vac tubes at a time. How long does burn in take? Of course, the answer is that it depends on the type, composition, and munufacturing process of the vacuum tube but usually varies from 10s to couple 100s of hours
What is burn in? The short answer is the removal of crud off the cathode.
The long answer can be elucidated by reading this excellent, but easily understood article linked below (all credit goes to the author’s):
Some excerpts from the above linked article:
What will burn-in do?
It seems virtually everything which can go wrong with a tube, will be less of a problem after burn in. It improves the cathode, the vacuum, the mica, cleans the metals, activates the getter, and even makes the glass at the inside anti static. Almost anything, you name it, and burn in will do it. The longer a tube was stored, the more it is going to benefit from burn in.
In short, burn-in starts these processes:
- Increase the emission, by evaporating contamination from the cathode surface.
- Make the cathode surface more homogeneous (emission wise), reducing noise.
- Make the tube curves overall better, reducing distortion.
- Clean the grids partially, for more stabile bias.
- Clean the mica, for less sputter or pop noise.
- Loosen too tight connections between mica and metal parts, for less microphonics.
- Improve the vacuum for less noise and closer factory specifications. (See Note3).
- Harsh sound will disappear, this is due to all the above effects together.
About the purpose burn in.
In tube production, it comes down to one thing, and that is keeping contamination away from the cathode. In a working tube, such contamination at first will be a positively charged ion in vacuum. As the working anode is positively charged,it pushes such ions away in all directions. Some land on the glass, recombine there with an electron, and in case the ion was not a gas, the resulting molecule will stay on the glass. Over time, the glass gets darker. Some other part will land on cathode, and compromise the emission. Some elememts don't do much damage, such as carbon, but others like Chromium are real emission killers. How to get rid of this? One way is designing the cathode a little bit too hot by default. That will slowly evaporate the contamination, but also wears out the cathode faster. Another process is to add emission enhancers to the Cathode coating. In the powder mixture, some small part of the Barium is always replaced by Aluminum and Strontium, because aluminium is a ideal emission starter of new born tubes, and Strontium generally enhances emission during the rest of the tube life. However that is not just doping it with fractions. In fact this requires so much Aluminum and Strontium, that the Barium content becomes
significantly lower. Yet, such tubes work initially better, and even tolerate a lot of contamination. They also have a much shorter formatting period (something like burn in, but not the same). On top of that, they need less burn in, and most of them will keep the desired specifications right from the beginning very well. All of this works very nice. Apart from one thing: There is substantially less Barium in the cathode mixture. However the Barium is the heart and soul of the tube. This is the substance which gets used up by simply using the tube, and there is no other way. Less Barium depot means less life time.
The purpose of burn in.
In short, this brings the new tubes closer to their final specifications. It should be clear, that only after
THIS CONDITION was reached, it makes sense to do the matching. Not somewhere half way, and for sure not match tube right out of the factory boxes.
The burn in effect has a tendency to disappear, most of all with low Barium content tubes. Though with such, due to the higher doping, it will restore again faster. This explains the observations of some tube doctors, demonstrating with a white doctor's coat on youtube, how they burn in tubes by over heating them with a normal tube tester. Telling how well this works. But is this really so....? Just do the math. Suppose he burns in 120 tubes, on his tube tester, and lets each tube run only 2 hours. Like this it takes him one day to burn in a quad, and it will take him a full month to do the 120 tubes and get 30 quads. So he will be tempted to compromise on the burn in time, which in this example was low already. And since such tubes burn in relatively fast, it's probably done like he says indeed.
When working with NOS and other more expensive tubes, these have generally a higher Barium content, so less Aluminum and Strontium, and lower cathode temperature. To make this possible, also higher base quality of the materials. Burn in of such tubes takes so much longer, nobody can seriously do this on a normal tube tester. It needs a device which can do many at a time. In the old days of tube production, the factories did 1000's at a time, filling complete factory floors. This picture is from the EI factory
a sort of self fulfilling prophecy tweak enough and it “works”
