Since this last circuit requires an internal coupling capacitor, placing the two tubes in a totem-pole arrangement is no longer necessary. Placing the two tubes in a more conventional cascading arrangement allows us to restore direct coupling between the two tubes.

     The above circuit shows an all-tube, constant current source loaded compound amplifier. The connection from the first stage to the second could have been made from the first stage's top triode's cathode, but that would have confused many into believing that the first stage comprised an SRPP amplifier. (If the top triode's cathode is not attached to an external load, the circuit would function as a nearly constant current source loaded grounded-cathode amplifier and since the 200-ohm resistor sees the same current flow as the 2k resistor, the signal present on the bottom plate must equal the signal on the top cathode. So all we have done is taken DC voltage shift upwards.)
     The two extra triodes opens up a rich vein of possible circuit modification. One direction we could move in is to reduce the complexity of the circuit. For example, we have two biasing sub-circuits: one for each constant current source section. If, on the other hand, we were to use the preexisting DC voltages within the circuit, we could simplify the circuit greatly. For example, the input stage's own cathode voltage can be used to bias the second stage's constant current source and the second stage's cathode follower's output can be used to bias the first stage's constant current source. The following circuit makes these points clear.

Compound amplifier with constant current sources and DC coupling between stages

     The schematic above shows how the constant current sources can be used with series resistors to limit the voltage across solid-state constant current sources. It also shows how we have returned to a modified version of the compound amplifier. Where can we go from here? Should we go anywhere else, as this circuit truly offers the maximum gain and the cleanest output. One move might be to replace the solid-state constant current sources with tube-based ones.
      The move to a tube constant current sources doubles the number of triodes needed. (This will not please many, as actual reason for using the SRPP circuit for many tube-circuit designers was not that it performs better than a conventional grounded-cathode amplifier, but that it allowed the designers to do something with the extra triode that most tube envelopes contain. Often this extra triode can be disabled by not heating its heater in 12.6-volt tubes, e.g. 12AU7, but not with 6.3-volt tubes. Psychology is often more important than engineering.)

< PREVIOUS   Copyright © 2002 GlassWare   All Rights Reserved