1. Paramour: Top Level Link.
             10% Distortion Tables vs Frequency. Output impedance vs Frequency.

2. Paramour: Stock Output Transformer, Stock  Choke, AVVT 2A3 M tubes,
             Right Amp, 10% distortion pictures, No distortion curves.
             This link opens a new window so you can compare it to other pages.

3. Paramour: Stock Output Transformer, EXO-04 Choke, AVVT 2A3 M tubes,
             Right Amp, 10% distortion pictures, distortion curves.

4. Paramour: Stock Output Transformer, Stock  Choke, Sino 2A3 tubes,
             Left Amp,  10% distortion pictures, distortion curves and many other pictures.

5. Paramour: Stock Output Transformer, Stock  Choke, AVVT 2A3 tubes,
             Left Amp,  10% distortion pictures, distortion curves.

6. Paramour: EXO-45-VS Output Transformer, Stock  Choke, AVVT 2A3 tubes,
             Left Amp,  10% distortion pictures, distortion curves,
             50Hz 100 Hz distortion curve comparison.

7. Paramour: EXO-45-VS Output Transformer, EXO-04 Choke, AVVT 2A3 tubes
             Right Amp, 10% distortion pictures, distortion curves.

The amp under test:

This is the left channel with the stock Paramour magnetics with a SINO 2A3 at 10 kHz, 2V/ div, 10 usec/ div, 10% distortion. (6.3 V rms)

This is the left channel with the stock Paramour magnetics with a SINO 2A3 at 1 kHz, 2V/ div, 100 usec/ div, 10% distortion. (6.27 V rms)

This is the left channel with the stock Paramour magnetics with a SINO 2A3 at 100 Hz, 2V/ div, 1 msec/ div, 10% distortion. (5.25 V rms)

This is the left channel with the stock Paramour magnetics with a SINO 2A3 at 50 Hz, 2V/ div, 2 msec/ div, 10% distortion. (3.81 V rms)

This is the left channel with the stock Paramour magnetics with a SINO 2A3 at 20 Hz, 2V/ div, 5 msec/ div, 10% distortion. (1.57 V rms)

Notes for Chart 2:
1. I took these distortion measurements with a Heathkit Sinewave generator that was sending 1.0 % distortion at 100 Hz and 1.5 % distortion at 50 Hz to the Paramour before the Paramour added any distortion of its own.
2. I could manually null the HP334A I was using better than the autonull would work.

Chart 2: Paramour with stock iron: Distortion vs Output Power into 8.45 ohms on 8 ohm tap with Sino 2A3 output tube. (Left Amplifier)

Lets look at the 10 kHz picture again:
This is the left channel with the stock Paramour magnetics with a SINO 2A3 at 10 kHz, 2V/ div, 10 usec/ div, 10% distortion, 8.45 ohm load on 8 ohm output tap. (6.3 V rms)

This is what the plate and grid are doing in this while the output is putting out 6.3V into an 8.45 ohm load.

With 6.3V into a 8.45 ohm load on the 8 ohm tap, this is what the plate and 8 ohm tap look like. The output voltage lags the plate voltage because of leakage inductance in the transformer.

This is what happens when you remove the 8.45 ohm load on the output. The output voltage goes up to 10V rms at 10 kHz.

We can see from the two  pictures above that we can get the voltage out of the magnetics at 10 kHz, we just can't deliver the current.
(I bumped the grid's offset voltage when I adjusted the time scale. That caused the GND to for the grid to move.)

At 10V no load output 10 kHz, this is what the plate and 8 ohm output tap looks like:

The phase lag in the output voltage goes away because the impedance of the leakage inductance is now small compared to the no load impedance on the output.

Lets look at the 50 Hz picture again:

This is the left channel with the stock Paramour magnetics with a SINO 2A3 at 50 Hz, 2V/ div, 2 msec/ div, 10% distortion, 8.45 ohm load on 8 ohm tap. (3.81 V rms)

Here is the plate and grid voltage for the 50 Hz 3.81 V rms output into 8.45 ohm waveform above.

Here is the plate and output voltage for the picture above. (50 Hz, 3.81V into 8.45 ohms)
Note that I changed the gain of the output from 2V/ div down to 5V/ div.

Now just for kicks, lets keep the grid voltage the same and pull the 8.45 ohm load off the 8 ohm output.
50 Hz with 5.38V rms output on the unloaded 8 ohm output tap:

That is some serious distortion. It looks like we are running out of plate current at the same time we are saturating the output transformer.