I designed a stereo 45 amplifier. The power transformer had only a B+ winding and one 6.3V winding, nothing else.
With series resistors from each lead of the 6.3V, the paralleled 45 filaments got 2.5VAC.
But there was a 2 resistor virtual center tap to reduce the hum (fixed resistors instead of a potentiometer) that virtual center tap connected to the shared self bias resistor and bypass cap, the other end of that self bias RC was grounded.
But the virtual center tap resistors connected to the filaments, which effectively shared the varying plate currents of the separate L and R channels.
That caused there to be only -40dBc of channel separation from 20Hz to 20kHz.
No listeners ever noticed the reduced channel separation, they loved that amplifier and speakers I played at work.
If you do not believe me about channel separation, then just listen to your favorite vinyl setup . . .
If you set Vinyl playback up just right, you get 30 to 35dB separation at mid frequencies.
And you get perhaps as good as 20 dB at 5kHz, but worse at higher frequencies.
And you get perhaps 10 to 15 dB at low frequencies.
How did that sound?
I love that sound!
Worry is one of the worst enemies of Enjoying the Music!
Worried by the Miller Effect Capacitance? Then drive the grid with a low impedance signal source.
An average CD player might have a 100 Ohm output impedance, or at least as low as 1k Ohm output impedance.
How much capacitance does it take to roll off 100kHz by -3dB if the driving impedance is 1k Ohm?
1592 pF. What tube has up to 1592 pF Miller Effect Capacitance??????
And if you find some such tube, it will be rolled off by -1 dB @ 50 kHz (much better than my tweeters, and much better than my ear too).
With series resistors from each lead of the 6.3V, the paralleled 45 filaments got 2.5VAC.
But there was a 2 resistor virtual center tap to reduce the hum (fixed resistors instead of a potentiometer) that virtual center tap connected to the shared self bias resistor and bypass cap, the other end of that self bias RC was grounded.
But the virtual center tap resistors connected to the filaments, which effectively shared the varying plate currents of the separate L and R channels.
That caused there to be only -40dBc of channel separation from 20Hz to 20kHz.
No listeners ever noticed the reduced channel separation, they loved that amplifier and speakers I played at work.
If you do not believe me about channel separation, then just listen to your favorite vinyl setup . . .
If you set Vinyl playback up just right, you get 30 to 35dB separation at mid frequencies.
And you get perhaps as good as 20 dB at 5kHz, but worse at higher frequencies.
And you get perhaps 10 to 15 dB at low frequencies.
How did that sound?
I love that sound!
Worry is one of the worst enemies of Enjoying the Music!
Worried by the Miller Effect Capacitance? Then drive the grid with a low impedance signal source.
An average CD player might have a 100 Ohm output impedance, or at least as low as 1k Ohm output impedance.
How much capacitance does it take to roll off 100kHz by -3dB if the driving impedance is 1k Ohm?
1592 pF. What tube has up to 1592 pF Miller Effect Capacitance??????
And if you find some such tube, it will be rolled off by -1 dB @ 50 kHz (much better than my tweeters, and much better than my ear too).
Last edited:
I noticed that the ANK1 I tested had a lot more of the sweep tone coming out of the other channel on my bench test setup than my own design which goes as far has each channel having its own 10h choke, if this ANK1 can sound this good whilst less stereo separation, I think I was putting my efforts in the wrong places, quite humbling!
And come to think of it, one of my fav amps is the zen triode, which has a shared bias resistor between the power tubes like in the scenario you describe. And the vinyl example was exactly the natural analogy my mind gravitated to, two channels coming from a single stylus. at the end of the day, our ears pick stuff up from the same mass of air as well.
And come to think of it, one of my fav amps is the zen triode, which has a shared bias resistor between the power tubes like in the scenario you describe. And the vinyl example was exactly the natural analogy my mind gravitated to, two channels coming from a single stylus. at the end of the day, our ears pick stuff up from the same mass of air as well.
There is nothing wrong with good channel separation, low distortion, good frequency response, etc.
But playing a Vinyl or CD recording has to be viewed in the overall system.
Microphone(s), Recorder, Production house, Vinyl setup or CD player, amplifier, loudspeakers, room.
I generally do not worry about -1 dB frequency response.
But I have seen amplifiers with no global negative feedback, and that have 3 low frequency poles.
If the poles line up, we have -1dB, -1dB, -1dB = -3dB. That is just the amplifier, and not anything else in the system.
The complete System is what gives or does not give pleasing results.
But playing a Vinyl or CD recording has to be viewed in the overall system.
Microphone(s), Recorder, Production house, Vinyl setup or CD player, amplifier, loudspeakers, room.
I generally do not worry about -1 dB frequency response.
But I have seen amplifiers with no global negative feedback, and that have 3 low frequency poles.
If the poles line up, we have -1dB, -1dB, -1dB = -3dB. That is just the amplifier, and not anything else in the system.
The complete System is what gives or does not give pleasing results.