Reduced Filament Voltage
- Thread starter toobdood
- Start date
Help Support GroupDIY Audio Forum:
The 1-Watt heater in 12A_7-type tubes can support around 15mA of cathode emission safely and with long life.
A 12AU7 working at 12mA should not be under-volted much.
A 12AX7 (or 12AU7) working at 0.6mA can be undervolted a lot without much problem.
So it isn't so much about the tube as the application.
In tube-mikes, you don't need a lot of current, and you want to reduce total heat. In most such cases, using standard tubes, you can work well below 6.3V.
Grid emission can be a problem on high-Mu tubes working at full cathode heat and very sensitive grid circuits (capacitor mikes). Reducing grid temperature by reducing cathode heat may help. However most tube mikes favor medium-Mu tubes which do not have this problem so much. And specific cases of high grid current may just be stray cathode-stuff sputtered onto the grid; try another tube.
A 12AU7 working at 12mA should not be under-volted much.
A 12AX7 (or 12AU7) working at 0.6mA can be undervolted a lot without much problem.
So it isn't so much about the tube as the application.
In tube-mikes, you don't need a lot of current, and you want to reduce total heat. In most such cases, using standard tubes, you can work well below 6.3V.
Grid emission can be a problem on high-Mu tubes working at full cathode heat and very sensitive grid circuits (capacitor mikes). Reducing grid temperature by reducing cathode heat may help. However most tube mikes favor medium-Mu tubes which do not have this problem so much. And specific cases of high grid current may just be stray cathode-stuff sputtered onto the grid; try another tube.
Thanks PRR! I am happy to know that I can manipulate my favorite little tubes in this way and know where the limits are. It doesn't make sense to ask more current from a tube at lower voltages than what it was designed for at normal voltages.
In the circuit I ran the 12AU7 in it was passing 1ma of current at 100v. That's running a tube rated for 2 watts per plate at 0.1 per plate. This was while the tube was running with a reduced filament voltage of 9.3 volts.
I am not after high-performance clean, but greasy, fat, juicy goodness.
Peace
In the circuit I ran the 12AU7 in it was passing 1ma of current at 100v. That's running a tube rated for 2 watts per plate at 0.1 per plate. This was while the tube was running with a reduced filament voltage of 9.3 volts.
I am not after high-performance clean, but greasy, fat, juicy goodness.
Peace
Junction
Well-known member
I don't know if this translates to low power audio use, but in high powered FM transmitters using tubes like 4CX15000 or similar, I have had tremendous success with performance and extended life when running the filaments slightly lower, the spec called for 6.3V, I run them at around 6.1VDC (regulated) with no degradation in performance, but tremendously extended tube life. At the nominal 6.3V we would typically get on average 14,000 hours life, at 6.1V I have been getting on average 21,000 hours life. If we drop the filaments much lower. i.e. below 6.0 V we do see instant performance degradation, but not sure how the lower voltage affects life. The objective was to find the lowest possible filament voltage without affecting performance, the result being a very happy tube and very long life.
Michael
Michael
> I don't know if this translates to low power audio use
Not very well.
Any loose atoms inside the tube become positive ions and slam into the cathode with enough force to cause damage.
Below 500V, we can use Oxide Cathode, which is a soft "frosting" which emits electrons VERY easily, and has no hard limit on emission. The 15mA/Watt guide can be exceeded without great loss of life. But plate voltages over around 500V or 1000V average will bombard the cathode and destroy it.
Up to 5,000V (your 4CX15000 says 10KV), we can use Thoriated Tungsten. Tungsten wire doped with thorium; when hot, the thorium is a monatomic film of liquid clinging to the tungsten. Emission is much lower, which means higher heater power for similar current, but the thorium film is self-healing to a degree so it will stand hard bombardment. To get good performance, the filament temperature must be closely controlled: too hot and the tungsten and thorium evaporate, somewhat cooler and emission falls to useless levels.
You were able to trim-down the heat because your specific application did not need the full 5 Amp rating, or the 5A rating was conservative. And actually, your 6.1V is within specs for 4CX15000.
The performance and trade-off are quite different, Oxide versus Thoriated. You just can't quick-kill Oxide, and it is hard to find emission-limited conditions. OTOH, Thoriated emission sags quickly as heat is lowered, and going much lower may kill it quick. When run "right", life really is limited by cathode wear-out... that's very uncommon in little oxide-cathode tubes which usually go gassy or have a melt-down accident long before the oxide is used-up.
At even higher voltages we must use plain tungsten, which has to be run just under melting-point to be much good; heat is very critical.
For lurkers: 4CX15000 suggested conditions are 10,000V plate supply, 4.55 Amperes cathode current. You are supposed to give the filament 6.3V +/-0.3V at 164 Amperes. In FM service this gives 36,000 watts output, 9,000 Watts plate dissipation. While the 1,033 Watt filament seems like a lot of power, the ratio of filament power to output power is very much better than puny-volted 6L6 etc.
Not very well.
Any loose atoms inside the tube become positive ions and slam into the cathode with enough force to cause damage.
Below 500V, we can use Oxide Cathode, which is a soft "frosting" which emits electrons VERY easily, and has no hard limit on emission. The 15mA/Watt guide can be exceeded without great loss of life. But plate voltages over around 500V or 1000V average will bombard the cathode and destroy it.
Up to 5,000V (your 4CX15000 says 10KV), we can use Thoriated Tungsten. Tungsten wire doped with thorium; when hot, the thorium is a monatomic film of liquid clinging to the tungsten. Emission is much lower, which means higher heater power for similar current, but the thorium film is self-healing to a degree so it will stand hard bombardment. To get good performance, the filament temperature must be closely controlled: too hot and the tungsten and thorium evaporate, somewhat cooler and emission falls to useless levels.
You were able to trim-down the heat because your specific application did not need the full 5 Amp rating, or the 5A rating was conservative. And actually, your 6.1V is within specs for 4CX15000.
The performance and trade-off are quite different, Oxide versus Thoriated. You just can't quick-kill Oxide, and it is hard to find emission-limited conditions. OTOH, Thoriated emission sags quickly as heat is lowered, and going much lower may kill it quick. When run "right", life really is limited by cathode wear-out... that's very uncommon in little oxide-cathode tubes which usually go gassy or have a melt-down accident long before the oxide is used-up.
At even higher voltages we must use plain tungsten, which has to be run just under melting-point to be much good; heat is very critical.
For lurkers: 4CX15000 suggested conditions are 10,000V plate supply, 4.55 Amperes cathode current. You are supposed to give the filament 6.3V +/-0.3V at 164 Amperes. In FM service this gives 36,000 watts output, 9,000 Watts plate dissipation. While the 1,033 Watt filament seems like a lot of power, the ratio of filament power to output power is very much better than puny-volted 6L6 etc.
I'll read this whole thread when I have a moment. But I had a 12bh7 with LT=6.3-7VAC and I liked the sound. I will soon test it at the proper LT and see if it is much different. Sorry if this is an ear-elphant comment, but I thought I would throw it into the mix.
> And there are those of us foolish enough to dream of using broadcast tubes as audio amplifiers.
Yes, I was thinking how many Hammond 125ESEs to series-parallel to load-up one 4CX15000. (Hundreds.) Altho you could rig something with a large 13KV:440V pole-pig. And yes, I have enough power in my fusebox to heat the filament and plate (but maybe not two for stereo).
Many broadcast tubes are MADE for audio amplifiers.
* "The Svetlana™ 4CX5000A/8170 is designed for audio and radio frequency applications."
* Audio Frequency Power Amplifier or Modulator, Grid Driven, Class AB1
* Typical Operation (two tubes, sinusoidal waveform)
* DC plate voltage 5000 V
* Maximum signal plate current 4.40 A
* Plate output power 13.5 kW
The logical use is to modulate a 25KW AM transmitter.
Huh! This thoriated-tungsten tube has specific instructions about heater voltage:
At rated nominal filament voltage, the peak emission capability of the tube is many times that needed for communication service. A reduction in filament voltage will lower the filament temperature, and this reduction will substantially increase life expectancy. The correct value of filament voltage should be determined for the particular application. Svetlana recommends that the tube be operated at full nominal voltage for an initial stabilization period of 100 to 200 hours before any action is taken to operate at reduced voltage. The voltage should gradually be reduced until there is a slight degradation in performance—such as power output or distortion. The voltage should then be increased a few tenths of a Volt above the value where performance degradation was first noted. The operating point should be rechecked after 24 hours.
The 807 -IS- a broadcast tube. Who the heck needs that much brute power in any other service? The 807 is better known, in a different bottle, as 6L6.
811 triodes did a lot of medium-range ship-ship radio in WWII, and are darlings of the DHT audio community.
813 has good potential if you are not afraid of grid current. SE 813 with 6L6 driver can put 45 Watts into sweet low 5K impedance.
Yes, I was thinking how many Hammond 125ESEs to series-parallel to load-up one 4CX15000. (Hundreds.) Altho you could rig something with a large 13KV:440V pole-pig. And yes, I have enough power in my fusebox to heat the filament and plate (but maybe not two for stereo).
Many broadcast tubes are MADE for audio amplifiers.
* "The Svetlana™ 4CX5000A/8170 is designed for audio and radio frequency applications."
* Audio Frequency Power Amplifier or Modulator, Grid Driven, Class AB1
* Typical Operation (two tubes, sinusoidal waveform)
* DC plate voltage 5000 V
* Maximum signal plate current 4.40 A
* Plate output power 13.5 kW
The logical use is to modulate a 25KW AM transmitter.
Huh! This thoriated-tungsten tube has specific instructions about heater voltage:
At rated nominal filament voltage, the peak emission capability of the tube is many times that needed for communication service. A reduction in filament voltage will lower the filament temperature, and this reduction will substantially increase life expectancy. The correct value of filament voltage should be determined for the particular application. Svetlana recommends that the tube be operated at full nominal voltage for an initial stabilization period of 100 to 200 hours before any action is taken to operate at reduced voltage. The voltage should gradually be reduced until there is a slight degradation in performance—such as power output or distortion. The voltage should then be increased a few tenths of a Volt above the value where performance degradation was first noted. The operating point should be rechecked after 24 hours.
The 807 -IS- a broadcast tube. Who the heck needs that much brute power in any other service? The 807 is better known, in a different bottle, as 6L6.
811 triodes did a lot of medium-range ship-ship radio in WWII, and are darlings of the DHT audio community.
813 has good potential if you are not afraid of grid current. SE 813 with 6L6 driver can put 45 Watts into sweet low 5K impedance.
tubologic
Well-known member
- Joined
- Mar 29, 2009
- Messages
- 72
clintrubber said:
Very interesting topic. Running a tube with reduced filament voltage usually results in degraded electrical performances and is not recommended by tube manufacturers in their application sheets. Sonic performances in an audio circuit are unpredictable as there are too many parameters involved. You must try it for yourself. BUT, tube reliability and lifespan could be greatly improved by (slightly) reducing the filament voltage. Tube reliability was the main issue in the design of the ENIAC (the 1st all-electronic computer) who used nearly 18,000 tubes ! Presper Eckert (Chief Engineer,ENIAC computer project) went to RCA (probably the most advanced tube development company in the world at that time) to gather some statistical informations and here's what he found (excerpt from a 1988 interview with Pres Eckert):
"... they (RCA) said they didn't have an awful lot of experience at the kind of high reliability I needed,either,but they had built some instruments in which they had run the filaments of the tubes a little lower than normal voltage and used them at less voltage and current than they were normally used at. That was done,incidentally,for other reasons,it was done to get low input current into an amplifier. But they observed when they did this that they got greatly extended lives in these tubes... We (at Eniac) ran the tubes at something more than five percent,but less than ten percent less than the standard voltage that they were intended for. I think we ran them at 5.8 volts... this was to increase the emission life ... we were really probably ultra-conservative in how we used the tubes... " Of course,nobody today would ever consider building something with 18,000 tubes inside but extending the lifetime of our precious little (expensive) "bottles" by reducing filament voltage is still an interesting option today. Don't forget all this is very tube-type and application dependant: for exemple the (Telefunken) AC701 tube used in vintage mic's don't like to be under-heated as this would soon result in very serious performance degradation. This is why the original NEUMANN P.S.U's are usually designed to deliver 4.2 VDC at their socket outputs to compensate for cable/connector voltage drop and get precisely 4.0 VDC on the tube heater. In electronics you don't get something for nothing and there's a trade-off between tube reliability/lifespan and sonic performances. Choice is yours...
My gut says you've picked the last possible place to look for more performance. You've really exhausted all the other variables? I know I haven't. I can't say I've found any noticeable sonic difference with the variable filament voltage pieces I have around. I have a lot of stuff running slightly under spec at 5.95 VDC because it's convenient to use surplus regulated linear 5VDC supplies. The same pieces sound the same with 6.3 VAC on them. And 6.6V. I am extremely skeptical of any obvious changes heard on a 6.3 type in a preamp, line amp, or comp when within tolerance, yet those claims exist. I can't disprove it either; just seems like trying to count the blades of grass in the yard. So many other big variables, up to and including the AC coming out of the wall.
This is the fun for me: I've used all types of resistors: Carbon Composition, Carbon Film, Metal Film, Metal Oxide, and Wire Wound. I've used different caps: Poly films of various makes and models, orange-drops, paper-in-oil, old Sprague Black Beauties, and I KNOW I am NOT off my nut knowing that I can hear differences in all of these components in a guitar amp or hi-fi. I've played with the plate current and voltage in and out of bounds to know how they can affect tone and this is one more of the MANY variables that one can play with: So why not?
Similar threads
Latest posts
-
-
Balancing card (2x unbal-to-bal & 2x bal-to-unbal) R vs G graph added- Latest: MidnightArrakis
-
-
