Hello,
I have heard of the merits of using AC on the filaments of the big triode power tubes. I am not sure I understand why this is to sound better but I will take it at face value at this point. The problem is the frequency of the AC is within the audible range at 60hz.....why not lower the frequency to something like 15hz? I am not sure how this could be done. I was wondering if anybody has ever toyed with this idea or ever achieved it. Would like to hear others thoughts or experience with this.
Jeff
I have heard of the merits of using AC on the filaments of the big triode power tubes. I am not sure I understand why this is to sound better but I will take it at face value at this point. The problem is the frequency of the AC is within the audible range at 60hz.....why not lower the frequency to something like 15hz? I am not sure how this could be done. I was wondering if anybody has ever toyed with this idea or ever achieved it. Would like to hear others thoughts or experience with this.
Jeff
There are issues I have read regarding too low of a frequency causing BTU heating fluctuations, but I'll leave that up to the experts. In my mind, producing 15 Hz is no small task; iron must be large if you want to transform it. PWM would be an option, but that gets quite complex and then you have to deal with high frequency hash.
If you're unhappy with 60Hz heating, I would suggest either DC constant current or high frequency quasi resonant type supplies at >40kHz.
FWIW, my AC heated projects have such minimal hum (1mV or less in push pull designs - SET is in the works) that I don't see any big issue with it. My latest build was a 6n6p PP into a 46 DHT PP into a 300B DHT PP. No hum pots, just one 'null' pot in a key location, and I get inaudibility.
If you're unhappy with 60Hz heating, I would suggest either DC constant current or high frequency quasi resonant type supplies at >40kHz.
FWIW, my AC heated projects have such minimal hum (1mV or less in push pull designs - SET is in the works) that I don't see any big issue with it. My latest build was a 6n6p PP into a 46 DHT PP into a 300B DHT PP. No hum pots, just one 'null' pot in a key location, and I get inaudibility.
Forget about cost, several problems emerge.
First the size of the iron goes up substantially.
Second, there has to be some means of generating the lower frequency.
The phone company used to do this by "passive" means, this may be worthy of some research and consideration. They used the lower frequency AC for "ring"!
One of the issues with DC apparently, or so I have been told is that one side of the filament can be hot and the other cool... I dunno.
Making a simple power oscillator that runs at 15hz surely can't be too difficult, but maybe a little lower is better if you have really good subwoofers? The good part is that frequency stability is not a big factor in this application, so if it drifts, no one cares!
_-_-bear
First the size of the iron goes up substantially.
Second, there has to be some means of generating the lower frequency.
The phone company used to do this by "passive" means, this may be worthy of some research and consideration. They used the lower frequency AC for "ring"!
One of the issues with DC apparently, or so I have been told is that one side of the filament can be hot and the other cool... I dunno.
Making a simple power oscillator that runs at 15hz surely can't be too difficult, but maybe a little lower is better if you have really good subwoofers?
The good part is that frequency stability is not a big factor in this application, so if it drifts, no one cares!_-_-bear
I think the issue with DC filaments on a DHT is that the cathode varies in voltage from one end to the other over whatever the DC supply is. Some tubes filaments are "center tapped" for this reason. I'm not sure how much a few volts really matters compared to the 100s of volts potential between the cathode and the plate. Maybe some interaction with the proximity of the grid.
As far as running at 15Hz...I don't think it will help all that much because it is often the intermodulation products that you end up fighting.
As far as running at 15Hz...I don't think it will help all that much because it is often the intermodulation products that you end up fighting.
Lower frequency --> higher transformer core area needed to prevent transformer saturation --> higher cost. You'd be better off going with a higher (super-sonic) filament frequency, though, it'll couple more strongly into your circuit through the Cgk and Cak.
I'm mulling over a 300B design that I've been toying with in my head for a while. Theoretically, the electron emission will be slightly higher in the end of the filament/cathode where the |Vgk| is lowest. As others have said already, I doubt it makes a huge difference, but I have considered using a +/-2.5 V DC supply rather than a single +5 V. I haven't made any conclusions yet. The 300B does not have a center tapped filament, so it's probably a moot point anyway. But that's a thought.
~Tom
I'm mulling over a 300B design that I've been toying with in my head for a while. Theoretically, the electron emission will be slightly higher in the end of the filament/cathode where the |Vgk| is lowest. As others have said already, I doubt it makes a huge difference, but I have considered using a +/-2.5 V DC supply rather than a single +5 V. I haven't made any conclusions yet. The 300B does not have a center tapped filament, so it's probably a moot point anyway. But that's a thought.
~Tom
I thought one or more of the Shuguang 300Bs did. HF is another option, but you still have to deal with IMD products.
maybe I don't understand something that is obvious to me. When I look at an old style light bulb the filament in that light bulb is hot all the way across. The voltage at one end is 117volts and progressively gets to 0 at the other end. Yet the lightbulb glows evenly across the whole filament. Is the same not true for tubes? Why would one end of the filament be colder just because the current is AC or DC. Common now. Reverse the dc current and voila the tube still glows. If one end of the tube is hot now the other end should be hot if you reverse the DC. Yeah right! The heat in the cathode filament would only vary if the device was poorly designed and not if the current was AC or DC. Early tubes were designed to run on Batteries for the simple reason most people did not have any electricty in their homes in the early days. When houses were wired up, AC became the common method to heat up tubes. less parts needed = cheaper costs to build. Early DC rectifiers were large and not reliable. There were millions of table radios sold that were called AC / DC They could run on 110 AC or 110 DC. The heaters were all wired in series and added up to 110 volts.
Thanks for all the insight. I just was wondering if this idea has ever been approached. I certainly understand that transformers working at that frequency would be huge. I was thinking of using a DC source with some sort of occilator that reversed the voltage at a frequency of 15hz or something like that. I am in the process of starting a 300B project that originally calls for AC filament heaters. I saw the DC filament solution built into the T-rex amp and was thinking of trying to duplicate it on my build. I doubt it will make a huge difference either way, my hearing is pretty poor.....I don't know why I am still so into this stuff considering I am half-deaf anyways....
Thanks for all the remarks.
Jeff
Thanks for all the remarks.
Jeff
This might be a stretch... (and i mean a big stretch). but power and audio waves inside the circuit are essentially the same. you (collectively speaking of you and the forum) might be able to concoct a phase change circuit with inverting and non inverting inputs. if the tube has a center tap and say its oh... take the 12at7 for example. you can feed it +/- 6 volts VAC and invert the signal(power) on one side. to me theoretically it would be the same as noise cancelling headphones. But for the trouble you might as well go DC
There's some merit in using (smoothed) DC instead of AC for a directly heated filament. Ideally, centre tapping the filament should reduce the effects of potential difference along the filament, but it's just as easy to centre tap the filament voltage source.
If the tube is indirectly heated, any hum is going to come from capacitive coupling between the filament and cathode. Again, DC or centre tapping should be just as effective.
I certainly wouldn't bother trying to use a 15 Hz AC filament source. It's much easier to use smoothed DC, center tapped. Claims that AC "sounds better" than DC should be examined with very carefully for supporting evidence.
If the tube is indirectly heated, any hum is going to come from capacitive coupling between the filament and cathode. Again, DC or centre tapping should be just as effective.
I certainly wouldn't bother trying to use a 15 Hz AC filament source. It's much easier to use smoothed DC, center tapped. Claims that AC "sounds better" than DC should be examined with very carefully for supporting evidence.
What are ears?
OT - Poor hearing doesn't have to eliminate the enjoyment of music. What are ears anyway. Mechanically, they have poor frequency response and as they degrade with age and abuse they lose the ability to consciously convey the higher frequencies so you can't "hear" above 14khz like me, or possibly worse.
But that is why they/we are so cool. They really are interpretive devices connected to a massive super computer that interpolates sound filling in expected blanks into pre-wired templates. Small subconscious elements of ambient signal are assessed via the hair cells and fluid movement and pressure of our ears, sinuses, skin and who knows what else to create an ambiance that we appreciate as music.
So, although it is disappointing to our analytical measuring external value system that we can "hear" so well, we still can get it, appreciate it and love the experience music has to offer.
Just enjoy, hearing is only part of what our brain does with sound.
Sorry, just the rambling of a neurologist...
OT - Poor hearing doesn't have to eliminate the enjoyment of music. What are ears anyway. Mechanically, they have poor frequency response and as they degrade with age and abuse they lose the ability to consciously convey the higher frequencies so you can't "hear" above 14khz like me, or possibly worse.
But that is why they/we are so cool. They really are interpretive devices connected to a massive super computer that interpolates sound filling in expected blanks into pre-wired templates. Small subconscious elements of ambient signal are assessed via the hair cells and fluid movement and pressure of our ears, sinuses, skin and who knows what else to create an ambiance that we appreciate as music.
So, although it is disappointing to our analytical measuring external value system that we can "hear" so well, we still can get it, appreciate it and love the experience music has to offer.
Just enjoy, hearing is only part of what our brain does with sound.
Sorry, just the rambling of a neurologist...
Rkinze
yes the voltage varies across the heater element so does it also in a resistor or an inductor or the current thru a capacitor. What if the heater is floating and not connected to ground then from where to where is the voltage different. A poorly built tube may have hot spots or cooler spots in its heater, the cathode could be unevenly coated, the grid wires could be unevenly spaced, the plate could be a bit warped, there could be drafts in the room causing the tube to be a bit cooler on one side than the other and on an on . What I was trying to say was DC will not heat the tube differently from 60 HZ AC. 15 Hz would not work. I am not sure what the delay time of the heater is but a lot of people can detect 50hz flicker so 15 hz may cause a slight temperature fluctuation which may cause cathode current to vary a touch and be picked up by the circuits.
yes the voltage varies across the heater element so does it also in a resistor or an inductor or the current thru a capacitor. What if the heater is floating and not connected to ground then from where to where is the voltage different. A poorly built tube may have hot spots or cooler spots in its heater, the cathode could be unevenly coated, the grid wires could be unevenly spaced, the plate could be a bit warped, there could be drafts in the room causing the tube to be a bit cooler on one side than the other and on an on . What I was trying to say was DC will not heat the tube differently from 60 HZ AC. 15 Hz would not work. I am not sure what the delay time of the heater is but a lot of people can detect 50hz flicker so 15 hz may cause a slight temperature fluctuation which may cause cathode current to vary a touch and be picked up by the circuits.
Yes, I am not arguing this fact. The element is heated evenly. It's not a question of emissions. It is a question of the cathode voltage at any particular point. I assumed you were asking about directly heated cathodes. Maybe I was wrong. For the DHT case, the difference is the voltage potential from one end of the cathode to the other relative to the grid, plate etc. It actually doesn't matter if you float the heater supply, there is still a difference of a few volts. Also keep in mind that you can't float the heater supply on a DHT! It's the cathode and must be grounded or held to some fixed potential relative to the plate/grid.
As far as indirectly heated cathodes, I don't see any disadvantages to DC heaters other than the inconvenience.
The element is heated evenly. It's not a question of emissions. It is a question of the cathode voltage at any particular point. I assumed you were asking about directly heated cathodes. Maybe I was wrong. For the DHT case, the difference is the voltage potential from one end of the cathode to the other relative to the grid, plate etc. It actually doesn't matter if you float the heater supply, there is still a difference of a few volts. Also keep in mind that you can't float the heater supply on a DHT! It's the cathode and must be grounded or held to some fixed potential relative to the plate/grid.As far as indirectly heated cathodes, I don't see any disadvantages to DC heaters other than the inconvenience.
picture a 5vdc 1-2 amp power supply. The dc power supply output is not connected in any way to any ground, ac earth or any electronic device. Now connect one end of this dc power supply to pin 1 one (cathode pin) and the other output of the power supply to the other cathode pin 2. If you use a cathode resistor to help bias the tube the dc power supply is "floating above ground". Check out this schematic they use a similar method. notice the balance pot across the 300b cathode. You will need to use one dc power supply per 300b tube
6SN7 / 300B Single-Ended (SE) Tube Amplifier Schematic
6SN7 / 300B Single-Ended (SE) Tube Amplifier Schematic
Using a low frequency will turn a hum problem into a tremelo problem. Could be very interesting for guitar effects; hi-fi, not so much. Nothing beats proper design (like avoiding DHTs, ohhh! 
).
If you've tried all the other approaches and still can't eliminate the hum, then DC or HF heaters is the way to go. 100kHz is easy to generate, usable directly (all the IMD products will end up modulating the 100kHz, which won't even get through the OPT), and easy to rectify if you'd like a more compact DC source.
Tim
).If you've tried all the other approaches and still can't eliminate the hum, then DC or HF heaters is the way to go. 100kHz is easy to generate, usable directly (all the IMD products will end up modulating the 100kHz, which won't even get through the OPT), and easy to rectify if you'd like a more compact DC source.
Tim
Last edited:
Sorry...doesn't change the fact that the DC potential from one end of the filament to the other is not the same. That balance pot is to try to cancel out 60Hz hum introduced by any unevenness in the AC filament supply. You want the DC potential of the cathode relative to the rest of the circuit "centered" across that AC supply as much as possible. The only reason it is floating is so that no cathode current can flow to the plate from that filament supply.
The filament supply, AC or DC, isn't really floating once the circuit is in operation. It is being held at a fixed voltage due to the cathode resistor. Take the schematic in the example you gave. Let's say the tube is biases itself at 60mA (wild guess). That puts 52V across the resistor. Now you also have 5V from your DC heater supply across that pot...100 ohms from its perspective. Now I am doing the circuit analysis in my head, but the DC going through the pot from the heater supply will actually unbalance the pot such that when you factor in the cathode currents, more current flows through one half than the other. The point is that your "floating" DC supply is being fixed around 52V at the top of the cathode resistor. If the supply keeps putting out 5V under these conditions, one end of the cathode is at one potential and the other is at a different potential.
Maybe I'm just dense, but that is how I see it.
The filament supply, AC or DC, isn't really floating once the circuit is in operation. It is being held at a fixed voltage due to the cathode resistor. Take the schematic in the example you gave. Let's say the tube is biases itself at 60mA (wild guess). That puts 52V across the resistor. Now you also have 5V from your DC heater supply across that pot...100 ohms from its perspective. Now I am doing the circuit analysis in my head, but the DC going through the pot from the heater supply will actually unbalance the pot such that when you factor in the cathode currents, more current flows through one half than the other. The point is that your "floating" DC supply is being fixed around 52V at the top of the cathode resistor. If the supply keeps putting out 5V under these conditions, one end of the cathode is at one potential and the other is at a different potential.
Maybe I'm just dense, but that is how I see it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.