|
In a normal pentode amplifier, grid 2 sees a constant DC voltage. In a triode connected amplifier, grid 2 sees the same DC and varying voltage that the plate sees, as this grid is connected to the plate (usually through a small valued resistor, 1k). The ultra-linear amplifier works between these two extremes. The method is simple: a portion of the plate voltage's swing is fed back into grid 2 via a low impedance tap of the output transformer's primary winding. This voltage represents a signal that is negative or anti-phase to input grid's signal. Thus, the output impedance at the tube's plate will decrease and so too its gain and distortion. But as the DC center voltage on grid 2 is at the same potential as the plate, the tube will share much of the pentode's higher potential power output, as the greatly positive voltage on this grid will make for much greater potential conduction at low plate voltages.
Ultra-linear operation of a tetrode or pentode, however, was said to offer much more than just a compromise between these two extremes. It offered an output impedance close to that of the triode connected pentode, a maximum power output equal or even greater than the strict pentode operation of the tube, a low level distortion figure almost as low as the triode, and a high level distortion figure lower than the pentode. The best of all possible output stages, argued Hafler and Keroes.
Well, maybe.
Although popular with tube amplifier manufacturers because of the higher power output it yields, Ultra-linear mode does not command a large avid following the way that triodes do. But then how many of us have heard a single-ended, directly heated pentode (e.g. 833), ultra-linear amplifier or a push-pull ultra-linear amplifier without feedback? Furthermore, many of the ultra-linear amplifiers we have heard may have suffered from the limitations imposed by the transformer. I have tested a few transformers whose ultra-linear taps seemed poorly tapped or whose output yielded a
|
|
