The output stage used in the Hi-End Hybrid Amplifier End 2023 and all the previous releases have an incredible sound much better than normal and common double or triple darlington.
Advantages of this type of configuration compared to a conventional:
● compensation of Vbe non-linearities between driver and final so very low distortion without feedback
● high slew-rate and large frequency band
● temperature run-away compensation without sensors if driver and final transistors share the same heatsink
● high input impedance if driver stage use current generator (I1 & I2) auto protected by short circuit on output
Disadvantages of this type of configuration compared to a conventional:
● the bias current in the driver limits the current supplied to the load
This design as been used in commercial products and the first is the Audio Research D100.Many people ask me a solution to reduce the power dissipation on driver and also in the commercial products some solutions as been used.
I have simulated and tested all these solutions but I noticed a worsening of the sound so I abandoned it.
Here I will show all the solutions with my test published also on diyaudio forum.
PIONEER A-09
This is the solution used in the Pioneer A-09, an output transistor with reduced power is used to detect the output current necessary to drive the load and its current is copied slightly amplified with a current mirror to modify the previous stage.
Many years ago I started testing this configuration but then abandoned it because there were oscillation problems that can only be resolved by significantly lowering the input resistance from 27K to 2k7.
This means that to drive the Pioneer A-09 final stage with a vacuum tube like my is necessary to add a buffer (see photo).
Here is the comparison between the 2 output stages D100 and A-09 and the better performance of the A-09 is evident.
The Left channel is the input of the output stage so the vacuum tube stage and the Right is the output.
On 4 ohm load the current on the driver of D100 is near to the end and therefore the distortion increases.
- Original schematic
- File to produce the pcb
The frequency response has been limited with an input filter at 500KHz becasue otherwise there was oscillation in the frequency response measurements and now have a limit 238kHz at -3dB.
A09 output stage + diamond buffer used (DC servo is not included)
The my Pioneer A-09 output stage sounds much worse than the original AR D100 configuration but I have tested low bias on output transistor from 5 to 100mA .
The condition of original Pioneer A-09 was 170mV on 0.44ohm so 0.386A for each final transistor but it means 0.386 * 4 = 1.54A.
The original amp. was powered with only 28+28 so 56VDC so 1.54A * 56V = 86W per channel.
AUDIO RESEARCH D400
This is the solution used in the Audio Reaserach D400, the output voltage is detect with a pair of mosfet and this modify the current generator bias.
Detect voltage instead of current is not correct because a correction is applied to the generators when it is not yet needed.
In this case 3 trimmer has been used:
1) level of the correction
2) bias of final transistors
3) balance the current generator to compensate the mosfet P/N differences
Only few opamp are good for this use, I have used SA5230 like original and 2SK216/2SJ79.
In this case the output stage pcb is the my original so you can add a module above.
Follows the distortion of only the output stage, on 4ohm we have a good working as expected.
