inductor. (Because the wire used in the making of an inductor has its own resistance, all inductors fail to realize their ideal. Still this DCR can be used in this arrangement by subtracting the resistance from the value of resistor Rak.)
   The case of the capacitive load is more difficult. We cannot just place a capacitor in series with resistor Rak, as no current will flow at idle. The solution is to replace Rak with a constant current source and to place a series capacitor-resistor combination in parallel with this current source. Thus at DC all the current will flow from the current source into the top triode. Since at DC the capacitive load will equal infinite impedance, it is fitting that the effective value of "Rak" also equals infinity. At the other extreme, infinite AC, the top triode sees only the resistor's impedance between it and the bottom triode's plate.

Triple Triode Mu Follower
   Many a tube enthusiast sees the SRPP as being an SE constant current source loaded Grounded Cathode amplifier that then cascades into an SE Cathode Follower that works into a constant current source load (see this month's letter section). In this view the gain must equal the mu of the triode and the output impedance must equal rp divided by the mu. Of course, the gain falls short of the mu by quite a bit and the output impedance is only about half of the rp. All of which makes sense, as only a resistor spans between the two triodes; and as a resistor is not a constant current source, it cannot truly isolate the two triodes as a true constant current source would. Why not replace the resistor with a constant current source then?

    This logic has led to the creation of the Beta-Follower. This hybrid circuit uses a transistor in place of the resistor. And for the most part it works. Unfortunately, the problems of accurately biasing up the transistor and the shunting of the current source because of the bias circuit are always troublesome.

At any intermediate frequency between DC and infinity, the effective value of "Rak" must satisfy this equation:
    Rak = rp + 2RL
                 mu - 1

For a capacitive load, the value of Rak is given by rp/(mu + 1) and we find the value of the Cak capacitor by:
    Cak = (mu - 1)C / 2,
where C is the capacitance of the load.
For a inductive load, the value of Rak is also given by rp/(mu + 1) and we find the value of the Lak inductor by:
    Lak = 2L  / (mu - 1).
where L is the inductance of the load.
   In theory, it is even possible to  compensate a very complex impedance such as a loudspeaker's by placing all the inverse reactive element together in a complex network between the top triode and the bottom triode.

    A better solution is to use a FET or depletion mode MOSFET as these devices allow using just a source resistor to set their idle current in much the same way a cathode resistor set the bias for a triode. The limitation to these devices is that they require a larger operating voltage than the transistor. This may not prove to be a problem when low rp, low mu triodes are used, such as the 6BL7, 6BX7, 12B4, and triode connected 6BQ5 and EL86, as these tubes require a large cathode-to-grid voltage to set their idle current.
    When the resistor that spans the bottom triode's plate to the top triode's cathode is replaced by a constant current source, the circuit