The circuit at the right shows a simple current-feedback power buffer. The input transistors provide some current gain and some voltage shifting for the following transistors, which can be regarded as making up the true input stage, as they are in charge of controlling the output. The emitters monitor the output and any deviation from the input will provoke a countervailing swing of the output stage's output. The following pair of transistors provide more voltage and current gain to drive the output transistors or MOSFETs. They also perform the needed phase inversion to bring the output in phase with the input.
    Translating the current feedback topology into vacuum tubes is a topic I do not think you will see in rec.tubes (or anywhere else). In the circuit below, the aims of current feedback have been achieved. The distortion is low and the bandwidth wide. The input cathode follower presents little capacitance to bog down the attenuator's output impedance. The 5687 input triode's high idle current easily cuts through the 6072 grounded-cathode amplifier's Miller effect capacitance. Lastly, the 5687-based cathode follower yields a low output impedance capable of driving severe loads and the feedback loop.
    Some are wondering why the tubes are in the wrong order. Why isn't the order: 6072, 6072, 5687, 5687? While such an order appeals to our

sense of ever increasing strength, it does not provide as great a bandwidth, as the 6072's high output impedance and low idle current cannot drive a 5687-based grounded-cathode amplifier as far as the reverse order. (Some tube designers like to move from the smallest tube to the largest, with zero thought to the consequences.)

Current-feedback amplifier that uses tubes. Definitely not your fathers line stage.

< PREVIOUS

www.tubecad.com   Copyright © 2002 GlassWare   All Rights Reserved

NEXT >

Pg.

14