Hi I was surprised when I looked at the Ongaku schematic to see the output is on AC. The Ongaku is known for is inky black soundstage and holographic imaging. Has anyone had experience with ac on 211se full power both chinese and nos tubes? Does the Ongaku suppress hum in the circuit? I tried 845 @450 and it was too noisy. Thanks for the help
AC filaments with a triode that is built somewhat like a 45, 2A3 (double filament/double plate), 300B, 845, 211, etc. . . .
1. How exactly is the filament centered and spaced between the 2 sides of the plate (Iron or steel plates, not carbon plates)?
AC filaments create an AC magnetic field. They form an AC electromagnet, causing the wires to move toward other magnetic materials.
Any uneven spacing of the filament to the two sides of the plate will cause the filament wires to move toward the closest side of the magnetic plate.
AC modulation of the plate current.
Hum.
2. How equal is the transconductance all the way from one end of the filament to the other end of the filament?
Any un-equal transconductance along that path will cause AC modulation of the filament to plate current.
That modulation causes 2 x Power Mains AC frequency to Inter-modulate with each and every musical tone, and each and every harmonic of the musical tones.
Two tone intermodulation (music tone or harmonic, and 2 x the power mains frequency). You need either a spectrum analyzer or FFT to see this.
Hearing is another matter, depending on the degree versus the tube you select; and your ears or you audience's ears.
3. The degree of the problems in # 1 and # 2 above, is a function of the exact tube that you use. No two tubes are exactly alike, which is why tube selection is very important to reduce # 1 above. # 2 above is harder to measure, but can be done, so there again, you are selecting tubes.
Now you have to select one that is best for # 1 and # 2.
4. Some give up on selecting so many tubes, and decide to power the filament on DC Voltage.
Choose wisely.
5. A hum balance resistor partially takes care of hum, but does not fix problems # 1 and # 2 above.
6. Your mileage may vary.
Mostly now, I use Indirectly Heated Beam Power, Pentode, and Triode tubes (uni-potential cathodes).
If I ever do a 2A3 amplifier again, it will have DC filaments this time.
Just my opinions and experience.
Happy designing; Happy building; Happy Listening.
I hope that gives you food for thought.
1. How exactly is the filament centered and spaced between the 2 sides of the plate (Iron or steel plates, not carbon plates)?
AC filaments create an AC magnetic field. They form an AC electromagnet, causing the wires to move toward other magnetic materials.
Any uneven spacing of the filament to the two sides of the plate will cause the filament wires to move toward the closest side of the magnetic plate.
AC modulation of the plate current.
Hum.
2. How equal is the transconductance all the way from one end of the filament to the other end of the filament?
Any un-equal transconductance along that path will cause AC modulation of the filament to plate current.
That modulation causes 2 x Power Mains AC frequency to Inter-modulate with each and every musical tone, and each and every harmonic of the musical tones.
Two tone intermodulation (music tone or harmonic, and 2 x the power mains frequency). You need either a spectrum analyzer or FFT to see this.
Hearing is another matter, depending on the degree versus the tube you select; and your ears or you audience's ears.
3. The degree of the problems in # 1 and # 2 above, is a function of the exact tube that you use. No two tubes are exactly alike, which is why tube selection is very important to reduce # 1 above. # 2 above is harder to measure, but can be done, so there again, you are selecting tubes.
Now you have to select one that is best for # 1 and # 2.
4. Some give up on selecting so many tubes, and decide to power the filament on DC Voltage.
Choose wisely.
5. A hum balance resistor partially takes care of hum, but does not fix problems # 1 and # 2 above.
6. Your mileage may vary.
Mostly now, I use Indirectly Heated Beam Power, Pentode, and Triode tubes (uni-potential cathodes).
If I ever do a 2A3 amplifier again, it will have DC filaments this time.
Just my opinions and experience.
Happy designing; Happy building; Happy Listening.
I hope that gives you food for thought.
Last edited:
The one I heard had a noticeable hum, but everything else was so good it kind of disappeared whilst the music was playing.
10 VAC filament output valves like 211 / 845 can be very quiet (limited by B+ ripple) if the filament winding is floating and the signal "cathode" connection is from a trimpot. Null without signal is very deep. As 6A3sUMMER has said, this null is only true at static (zero signal) conditions, and deteriorates with signal, effectively a modulation of signal by the heating voltage.
Two possible approaches to this issue are to operate in push-pull, letting the output transformer's common mode rejection work as it does with even-order harmonics from the output stage, or to operate two identical valves in parallel with filaments in opposite polarity, which likewise cancels even-order modulations, but only from the heating voltage.
For those with an Eimac jones, type 304TL has separate filament connections of each pair of internal triodes. You do the math.
All good fortune,
Chris
Two possible approaches to this issue are to operate in push-pull, letting the output transformer's common mode rejection work as it does with even-order harmonics from the output stage, or to operate two identical valves in parallel with filaments in opposite polarity, which likewise cancels even-order modulations, but only from the heating voltage.
For those with an Eimac jones, type 304TL has separate filament connections of each pair of internal triodes. You do the math.
All good fortune,
Chris
Well That explains why I thought it was DC! I have been dreaming about building a big tube amp for years and have been collecting the parts. A dual choked unregulated supply usually are my favorite because of the technical simplicity but is a big job. As the years' go by heat and energy costs matter more in my plans. My living rooms get smaller. Iam thinking about building an integrated Amp to save running all the preamps supply's. To compensate a little for the outputs. If I remember right Kondo used it in a small room with coaxials which is the same as my setup with 94db speakers. I don't think the ongaku was made to be a Amp for horns. I imagine it was built to be and is the best simple full range integrated hifi amp ever!. I do think possibly in a simple system like his the ac would have sounded better than a raw unregulated unchoked dc supply. Thank you everyone for your help. I will post when I build.
I built brute force 5V DC power supplies for 300B filaments.
Secondary, Schottky bridge, 22,000uF, 2 Ohm series Resistor, 22,000uF, 300B filament.
Unregulated.
The ripple was 1mVrms. The ripple was 71dBc versus 5VDC.
The power mains went from 117v to 123V.
The 2 Ohm resistor was adjusted to give 5V with the power mains at 120V (center voltage of 117V to 123V).
This is a soft start power supply (automatic: 2 Ohm series resistor versus 1 Ohm 300B cold filament, and 4 Ohm 300B hot filament).
The 5V was only +/- 2.5% due to the mains power variation.
WE 300B calls for no more than +/- 5% DC variation.
Who needs a regulator? . . . You do if your power mains vary more than +/- 5%.
Regulators work.
So does brute force.
Secondary, Schottky bridge, 22,000uF, 2 Ohm series Resistor, 22,000uF, 300B filament.
Unregulated.
The ripple was 1mVrms. The ripple was 71dBc versus 5VDC.
The power mains went from 117v to 123V.
The 2 Ohm resistor was adjusted to give 5V with the power mains at 120V (center voltage of 117V to 123V).
This is a soft start power supply (automatic: 2 Ohm series resistor versus 1 Ohm 300B cold filament, and 4 Ohm 300B hot filament).
The 5V was only +/- 2.5% due to the mains power variation.
WE 300B calls for no more than +/- 5% DC variation.
Who needs a regulator? . . . You do if your power mains vary more than +/- 5%.
Regulators work.
So does brute force.
Well I have a 211 valve amplifier (not built by me, but will be updated/modified on the power supply and heaters)
My amp has 11.2V AC heaters on each channel that after rectification full bridge goes through 47,000uF R82 and 47,000uF
I am going to try the Rod Coleman regulators, and then if the voltage is not sufficient to drive the regulators I will try an inductor to replace the resistor. These seem appropriate used in parallel configuration at a similar impedance to the R82.
I have also modelled briefly this on PSUD2 to try and make sense of the response and ripple. The simulation suggests 31mV of ripple, I will try and measure what I actually get. One day I will invest in an oscilloscope, but for now I cheat with REW software and it's a faff :-(
How relevant is a soft start for the 211 valve?
My amp has 11.2V AC heaters on each channel that after rectification full bridge goes through 47,000uF R82 and 47,000uF
I am going to try the Rod Coleman regulators, and then if the voltage is not sufficient to drive the regulators I will try an inductor to replace the resistor. These seem appropriate used in parallel configuration at a similar impedance to the R82.
I have also modelled briefly this on PSUD2 to try and make sense of the response and ripple. The simulation suggests 31mV of ripple, I will try and measure what I actually get. One day I will invest in an oscilloscope, but for now I cheat with REW software and it's a faff :-(
How relevant is a soft start for the 211 valve?