Just as in the October 1999 issue's analysis of the White Cathode Follower, we find that if we try to achieve a very high gain and a low output impedance, we end up limiting the amount of potential current swing the amplifier can swing into the load.
As an example, if we make Rak equal to 100k and use a 6DJ8 with a 1200 volt power supply and an idle current of 10 mA and 600 ohm load, the static specifications are stellar: the gain is 27.8  and the Zo is 79 ohms. But the maximum Class A negative output voltage swing is only 7 millivolts! Positively, the amplifier can swing 15 volts and still remain in Class A mode. Why the huge discrepancy?

Any variation in the current flowing through resistor Rak will result in a signal voltage for the top triode. But the top triode can only see so much signal voltage before it either stops conducting or is overdriven.
The test of the degree of balanced operation we achieve is extent to which the peak negative output voltage swing equals the peak positive  output voltage swing.
The idle current  divided by the effective transconductance of the top triode gives us the maximum negative peak voltage the top tube can see before it cuts off. And this peak voltage divided by Rak gives us the peak increase in current through Rak needed to turnoff the top tube. This quantity of peak increased current  through the load impedance defines the peak negative output voltage swing. Or expressed mathematically:
Gm“ =                           mu

 SRPP circuit with fantastic calculated  specifications and very poor actual performance

rp + [(mu + 1)Rak + rp][(Rak+rp)||Rl]
rp + Rak
Vpeak =    Iq / Gm“ + IqRak
Imax =      Vpeak / Rak
Vn max =  Rl x Imax.
(The formula for Gm“ assumes that the cathode resistor is bypassed on the bottom tube.)

 pg. 6