SE: DC-coupled 12AT7 shunt-cascoded and KT88 in UL+UNSET

[zintolo]

Hi, while waiting to rebuild my area to play with circuits, looking for simple circuits I came out with this:

The driver is a 12AT7 in shunt cascode mode, whilst the power amp is a KT88 with 23% UL and 25% UNSET.
They are DC coupled because the bias of the output tube is set by the trimmer U5.

Most probably it is not the best working point, as at 1 Wrms it already has 0.8% 2nd harmonic (while all others are at least 32 dB lower):

Harmonic    Frequency     Fourier     Normalized     Phase      Normalized
 Number       [Hz]       Component     Component    [degree]    Phase [deg]
    1       1.000e+03    4.015e+00    1.000e+00       -0.35°        0.00°
    2       2.000e+03    3.347e-02    8.338e-03      -90.97°      -90.62°
    3       3.000e+03    8.452e-04    2.105e-04        1.15°        1.51°
    4       4.000e+03    1.117e-05    2.782e-06        8.54°        8.89°
    5       5.000e+03    2.712e-06    6.756e-07       -2.77°       -2.41°
    6       6.000e+03    5.595e-06    1.394e-06        0.46°        0.81°
    7       7.000e+03    4.842e-06    1.206e-06        0.67°        1.02°
    8       8.000e+03    4.213e-06    1.049e-06       -2.11°       -1.76°
    9       9.000e+03    3.733e-06    9.300e-07        0.60°        0.96°
Total Harmonic Distortion: 0.834107%(0.834107%)

but it keeps the 2nd dominant until and above the 21.5 Wrms at 5% THD:

Harmonic    Frequency     Fourier     Normalized     Phase      Normalized
 Number       [Hz]       Component     Component    [degree]    Phase [deg]
    1       1.000e+03    1.852e+01    1.000e+00       -0.37°        0.00°
    2       2.000e+03    7.667e-01    4.139e-02      -91.65°      -91.28°
    3       3.000e+03    4.643e-01    2.507e-02        0.46°        0.83°
    4       4.000e+03    5.015e-02    2.707e-03       80.91°       81.28°
    5       5.000e+03    2.048e-01    1.105e-02      179.76°      180.13°
    6       6.000e+03    6.929e-02    3.740e-03      -94.03°      -93.66°
    7       7.000e+03    9.398e-02    5.073e-03       -1.66°       -1.29°
    8       8.000e+03    4.950e-02    2.672e-03       85.53°       85.90°
    9       9.000e+03    3.275e-02    1.768e-03      176.16°      176.53°
Total Harmonic Distortion: 5.020704%(5.025475%)

and has a Rout of 1.36 Ohm without using any feedback.

A question for @Tubelab_com : can I double the FQPF9P25 having a parallel source follower to share the 18W to be dissipated?
I will have twice the input capacitance, but it will handle the power in a safer way.
I will of course try also the suggested 10% a-g1 local feedback, because in SE there's no issue with high g2 voltages as I have in the PP designs.

 

[zintolo]

A different driver and better results:

With this configuration the 2nd harmonic predominant up to 22 Wrms and 2%THD, then 3rd increases up to reaching 5% at 27.5 Wrms.
More harmonic cancellation can be obtained working on R1.

 

[Bigun]

I'm no expert, but is the Zout of the Pre low-enough to be useful ? or is that just a convenience for simulations

 

[zintolo]

Hi bigun, FQP9P25 Crss is 27 pF, so -3dB point is 180 kHz.