As with EL34, 6L6, KT77, KT88 and 6550; most Hi-Fi amplifiers from 1955 to 1975 operate power tubes at or near maximum plate dissipation ratings.
For this very reason, tubes that test fine at lower currents and voltages may tend to run-away when operated at real-life levels of current and voltage. The proof is in the operation at real voltages and currents.
Cathode biased amps and some amps that employ Pentode connected output designs generally are easier on Power Tubes as the Cathode resistor and the Lower G2 Voltage tends to throttle the tubes current better, i.e. the design is less prone to tube related issues such as mis-aligned internal structures and systematic manufacturing defects.
Fixed bias amps almost 'always' require 'matched pairs' of power tubes as there is no feedback 'throttle control' from the Cathode resistor. Often when connected in Ultralinear mode, screens (G2) may end up operating at a voltage beyond the maximum recommended as not enough voltage drop is provided by the primary transformer windings between the B+ and the Ultralinear taps.
Also some amps are known to run tubes hard and quicly weed out the weaklings.
The true acid test for Power Tubes; install the power tubes in the amplifier, verify proper bias, and in a darkened room, verify that the plates are not turning a dark BBQ Charcoal Red color.
If you have the technical ability to measure high voltages you can take a VOM and measure Plate voltage and Plate current. Their multiplication Volts x Amps = Watts i.e. the plate dissipation with no music appiled. For a 6550 tube this value should be less than 35 Watts, EL34 less than 25 watts, 6L6..... (see tube manuals )according to specifications.
Other factors that contribute to a tubes early demise is exceeding maximum Cathode to Heater voltage differences and/or exceeding the dissipation and voltage rating of the control grid G2 in pentodes and tetrodes. For a 6550 tube is somewhere near 6 watts max.
Most 40% taps on ultralinear output transformers may not 'drop' enough voltage towards G2 so as to maintain the level of current and voltage below Grid dissipation. Yet some 6550 tubes appear to resist 'stress' well, while others simply turn red after 2 minutes, despite being 10% under maximum dissipation ratings in all parameters.
One can take a calculator and paper to most 1960; EL34, KT88 and 6550 amps and find that the quiescent current and voltages may place the dissipation of the plate at or very near maximum, this was common design practice at the time for maximum power and with hopes of minimum distortion.
Most legendary NOS tubes such as Genalex, Tung-Sol, GE, RCA, Telefunken, Amperex and Sylvania either work flawlessly or dim like a light bulb burning out in a few hours, new production tubes are also similar, they either work or they don't. There does not seem to be a middle ground here as with smaller 9 pin tubes.
Tubes that operate with slightly 'red' plates typically exhibit much greater distortion than those that visually indicate no red plates and dissipation ratings below maximum.
In conclusion, 'expect' most Power Tubes to have issues in most Hi-Fi amps. The majority, but not all Guitar amps tend to operate their output stages in Class B, offering greater power but increased distortion.
RAM Labs tests power tubes at varying voltage and current levels up to 450 Volts DC.
It may very well be that some tube vendors may offer Hi-Fi rated duets and quads tested close to their maximum ratings. This way one can be assured that a 51 milliamp 6550 will operate properly at 70 milliamps in a Dynaco MK3.
Happy Tubing!
