Hi 6A3sUMMER,
I agree with you on everything you said. AC filaments wouldn't work well in anything but the o/p stage. Other tubes have lower current requirements anyway.
Don't forge the low Gm of the output, AC hum will be very low unless you messed up the supplies.
I agree with you on everything you said. AC filaments wouldn't work well in anything but the o/p stage. Other tubes have lower current requirements anyway.
Don't forge the low Gm of the output, AC hum will be very low unless you messed up the supplies.
Complaining about a tube being old in a tube forum is laughable. Is there cutoff year for what's acceptable and what is not?
There's no reason you can't try AC filaments again and see what the results are. In my 845 builds, noise issues apart, I felt that DC supplies with a common mode choke at the end actually sounded better.
Years ago I build a single ended 845 amplifier with the power supply on a separate chassis. It uses AC heating with balance pots and the hum is only audible if you put your head next to the speaker. I enjoy the industrial beauty of these large "ancient" tubes.
The issue of signal modulation by AC filaments can be attacked by using even numbers of identical valves in parallel with opposite AC polarity or in push-pull with same AC polarity.
In type 304TL this is just a matter of wiring choice. Type 211 has such a demanding ideal load impedance that it begs to be used in parallel anyway.
All good fortune,
Chris
In type 304TL this is just a matter of wiring choice. Type 211 has such a demanding ideal load impedance that it begs to be used in parallel anyway.
All good fortune,
Chris
Well, I've been thinking/talking/writing of power tubes, of course. Never would I try to use directly heated tubes in small signal applications (except for battery tubes, when low power consumption is mandatory). Even tube technology did observe severe progressions during the decades.
Best regards!
2. AC on the filament; 60Hz power mains; with a flute playing a note, Concert A 440Hz
440Hz, 880 Hz, 1320Hz, but you get something more, because of AC heating: 2 X 60 Hz = 120 Hz.
You will get 120Hz sidebands on each an every note and on each and every harmonic.
440 - 120 = 320Hz; 440 + 120 = 560Hz
880 - 120 = 760Hz; 880 - 120 = 760Hz
1320 -120 = 1200Hz; 1320 - 120 = 1440Hz
Not very good
3. Well, suppose you have 50Hz power mains. The results are very similar.
50 x 2 = 100 Hz
So, 440 - 100 = 340Hz; 440 + 100 = 540Hz
880 +/- 100
1320 +/- 100
You get the idea
Not Good
If you use AC heating on an input tube, driver tube, output tube that is a DHT, you get the same additional tones.
Have you ever played an amplifier so loud that it is very near to drawing grid current.
With AC DHT, it will draw grid current, at 2 x the power mains frequency.
The sidebands that appear below and above each and every musical tone, and each and every musical harmonic (overtone) will be very loud (120Hz or 100Hz)
Not Good
Just my opinions
Something wrong here.....
Since this is AC and NOT bridge rectified, you don´t get the double frequency (100 or 120 Hz).
You get exactly, what´s in the mains: 50 or 60 Hz hum. Btw easily cancelled out by using a "hum-bugger" (balancing pot on the filaments
to cancel hum), or you can use a hum cancelling feedback circuit
It's a nice suggestion, but as a check on how effective it's likely to be: consider the path of the AC waveform on the anode-cathode curves (and imagine it for the grid to cathode curves, which are also exercised by AC-heat). In both cases, the waveform tracks above and below the operating point; and as the music-signal amplitude increases, the (above→below) symmetry breaks down. This degrades the effectiveness of any cancellation.
Measurement of PP with 2 DHTs & same-polarity AC-heat still reveals large amounts of IMD...
Instead of lingering in AC, try an "ideal bridge". These 'rectifiers' are managed with a chip, and it ensures there are no glitches. The DC is great in structure. With your current bridge there will be some glitches, and even at -40/50 dB, they can be heard. It gives a hardness to sound. The special rectifiers (eg Rods, or Tent Labs that I both use) are a good way out.
OK. My experience is that AC a DHT heating has a warmth - specially if PP. Charm. But best to have a separate transformer. Otherwise it picks up the residue of the other windings.
Modulation by AC filaments depends on the valve's distortion - an ideal linear valve will have no modulation, but will sum linearly and can be trimmed out to zero. Modulation in a real triode is (largely) second order and decreases monotonically with signal level. Whether or not this is the optimum choice is a separate question, of course.
All good fortune,
Chris
Take your favorite SE amplifier that has a DHT output stage, and that has an AC powered DHT output tube filament.
Set the AC filament balance pot until you have 1mV or less of hum into a proper load resistor (if you can).
Apply a pure 1kHz tone at the input, and set the gain until the amplifier is putting out 1/2 of full power (the voltage will be 70% of the full power voltage, -3dB from full power).
Take a capable spectrum analyzer, scope FFT, or sound card and software.
Look at the 1kHz tone.
Question: What are those 2 extra tones that are either 100 or 120Hz below 1kHz, and 100 or 120Hz above 1kHz?
Those are the intermodulation products of AC filaments and a DHT.
100Hz for 50Hz power mains
120Hz for 60Hz power mains
Do you care about those 2 extra tones that were not originally on the test signal?
Do you care about those 2 extra tones that were not originally on the music signal?
IRMC
Set the AC filament balance pot until you have 1mV or less of hum into a proper load resistor (if you can).
Apply a pure 1kHz tone at the input, and set the gain until the amplifier is putting out 1/2 of full power (the voltage will be 70% of the full power voltage, -3dB from full power).
Take a capable spectrum analyzer, scope FFT, or sound card and software.
Look at the 1kHz tone.
Question: What are those 2 extra tones that are either 100 or 120Hz below 1kHz, and 100 or 120Hz above 1kHz?
Those are the intermodulation products of AC filaments and a DHT.
100Hz for 50Hz power mains
120Hz for 60Hz power mains
Do you care about those 2 extra tones that were not originally on the test signal?
Do you care about those 2 extra tones that were not originally on the music signal?
IRMC
Hey - thanks everyone for the help and thoughts so far.
My take on this at the moment is that DC is surely going to help on minimising hum, and AC comes with intrinsic challenges to get to a low hum amplifier (I have 99db/W speakers), but that said it is not impossible just needs really good execution.
DC may have some issues on long term life of tubes, BUT this is not a serious problem.
All DC heaters are not equal, and mine as CRC is a little primitive, and I may find some improvements (e.g. rectifiers) if I look to upgrade this supply.
So far I am not hearing a lot of votes that suggests AC is intrinsically better sounding but with the caveat of the hum challenges?
My take on this at the moment is that DC is surely going to help on minimising hum, and AC comes with intrinsic challenges to get to a low hum amplifier (I have 99db/W speakers), but that said it is not impossible just needs really good execution.
DC may have some issues on long term life of tubes, BUT this is not a serious problem.
All DC heaters are not equal, and mine as CRC is a little primitive, and I may find some improvements (e.g. rectifiers) if I look to upgrade this supply.
So far I am not hearing a lot of votes that suggests AC is intrinsically better sounding but with the caveat of the hum challenges?
That was my experience with indirectly heated filaments as well. Sidebands are easily visible and even though the increased hum wasn't audible at a distance the impact on the overal sound certainly was audible.
I don't run DHTs but it seems intuitive that good hum cancellation requires low harmonic content and high symmetry on AC filaments. Is that correct? It's been decades since clean and symmetrical 60Hz was a given from public utilities in my area. My provider sets a distortion limit of 8%.
No doubt. This was a single globe two stage pentode mode SE with plenty of global feedback, without which it would have been much worse. Both tubes operate close to dissipation limits which apparently increases susceptibility. No variation of AC filament including positive and negative voltage offset totally eliminated hum. For build simplicity alone I wish it did.