The easiest, and most reliable approach is to always have at least one new set of replacement tubes in reserve. Every year or so, replace the tubes with the new set and see if the sound has changed markedly for the better; if so, it is time to replace the tubes. Do this separately for small signal and power tubes because small signal tubes usually last MUCH longer.
Tube testers are generally NOT reliable for this. Most testers are good for spotting gross problems but not subtle deterioration. Most testers are not terribly reliable because they need to be calibrated every so often and who really sends them away for that kind of maintenance?
One of the best testers out there is one that is in current production and is quite expensive. It is the Amplitrex tester. It is pretty much idiot proof (a screen tells you exactly what to do step by step), accurate, free of need for recalibration and can test for a wide range of factors. You can even hook it up to a computer to trace curves. One nice feature is that it will give real, meaningful readouts (e.g., transconductance) and not some proprietary "value," provide the tubes specifications to compare with the test results, as well as provide gross evaluations ("excellent," "good," "weak" and "bad."). I have this tester and like it a lot. The only big weakness is the limited number of tubes that are programmed into the tester (you can do your own programming if you have the proper data or you can have the factory program a CD ROM to update your machine).
The problem with any tester is that, depending on the application, a tube may test extremely bad and still be perfectly fine for the particular use in the amplifier. I have a number of extremely weak 6sn7 tubes that sound WAY better than any current production tube in my amplifier. I doubt this would be the case if the tube were pushed much, but, in this amplifier the tube's capabilities are taxed very much. Even the Amplitrex is best used for testing to see if a tube is generally good, and the extent to which the vendor stretched the truth as to its "age," and not for making any kind of objective evaluation of the sound of the tube. My amplitrex has already proven itself by showing that a "new" rectifier should not be used (no voltage drop in EITHER direction, i.e., shorted out).
Tube testers are generally NOT reliable for this. Most testers are good for spotting gross problems but not subtle deterioration. Most testers are not terribly reliable because they need to be calibrated every so often and who really sends them away for that kind of maintenance?
One of the best testers out there is one that is in current production and is quite expensive. It is the Amplitrex tester. It is pretty much idiot proof (a screen tells you exactly what to do step by step), accurate, free of need for recalibration and can test for a wide range of factors. You can even hook it up to a computer to trace curves. One nice feature is that it will give real, meaningful readouts (e.g., transconductance) and not some proprietary "value," provide the tubes specifications to compare with the test results, as well as provide gross evaluations ("excellent," "good," "weak" and "bad."). I have this tester and like it a lot. The only big weakness is the limited number of tubes that are programmed into the tester (you can do your own programming if you have the proper data or you can have the factory program a CD ROM to update your machine).
The problem with any tester is that, depending on the application, a tube may test extremely bad and still be perfectly fine for the particular use in the amplifier. I have a number of extremely weak 6sn7 tubes that sound WAY better than any current production tube in my amplifier. I doubt this would be the case if the tube were pushed much, but, in this amplifier the tube's capabilities are taxed very much. Even the Amplitrex is best used for testing to see if a tube is generally good, and the extent to which the vendor stretched the truth as to its "age," and not for making any kind of objective evaluation of the sound of the tube. My amplitrex has already proven itself by showing that a "new" rectifier should not be used (no voltage drop in EITHER direction, i.e., shorted out).