Please comment on constant current for tube's filament using
Amperite Ballastic current regulators like 7H-4B etc.
You could find some helpful informations and discussions on certain website(http://tubeaudio.netfirms.com/top_UltimDIY.htm) concerning this topic.
I have not myself use them in my circuit so I haven't got the first hand experience of these tubes but somehow the informations I got are very positive!
Amperite Ballastic current regulators like 7H-4B etc.
You could find some helpful informations and discussions on certain website(http://tubeaudio.netfirms.com/top_UltimDIY.htm) concerning this topic.
I have not myself use them in my circuit so I haven't got the first hand experience of these tubes but somehow the informations I got are very positive!
Attachments
tubetvr said:
It was my conclusion to previous explanations, but if you are so curious here you go:
Constant current means high dynamic resistance (approaching infinite ideally), constant voltage means low dynamic resistance (approaching zero ideally). Resistance of a filament is proportional to its temperature. When tube is cold yet, filament's resistance is low. When tube is hot it filament's resistance is higher.
A filament is a spiral wound from a tungsten wire that can't be absolutely equal in diameter and turns can't be absolutely equal close to each other and to the cathode they heat. It means, when you apply a voltage to the heater some parts of it are getting hotter than others. It means their temperature is higher. It means, their resistance is higher, and it causes higher temperature. Local overheating happens that causes mechanical deformations and uneven vaporizing of filament material that is thinner and thinner each time you power your tube on.
When we limit inrush current the filament heats up slower, so temperature spreads more equally across it, that prolongs life of your precious tube.
No. Since filament resistance is equal to temperature when working with constant voltage temperature is more stable. It may be more stable with a regulator with negative output dynamic resistance. I used such approach in equipment I designed to calibrate a temperature measurement tool that senses axle bearing temperature in passing by train wagons.
Practical example?
Here you go:
Mikewong said:
That days making ANYTHING constant was The Big Achievement. Gas discharge tubes, rectifying tubes, regulating power tubes could not work effectively on filament voltages. Magnetic amplifiers were known, but they were heavy and expencive. Ferro-resonant stabilizers were heavy and expensive, also They required constant frequency. Constant frequency was expensive, and lot of applications were battery powered.
Bulbs could, they were cheap (probably, they will be expensive after fluorescent light sources will take over the World wiping out the technology), and they regulated current, that was better than to regulate nothing. Ballastic current regulators ARE electric bulbs rated strictly to regulate the current.
Today we have a choice, right? We are happier than engineers 70 years ago, we have much more toys to play with and choose from...
You obviously didn't read what I was writing earlier about heater failure NOT being the most common failure mode for tubes
No, it can maybe prolong the life of the heater but that doesn't have to be the same thing as prolonging the life of the tube.
And can you show that this has any practical effect of performance in electronic tubes?
BTW, the constant current regulator tube described in another posting and that you commented on consist of a iron wire in a hydrogen gas filled envelope, it is not an ordinary light bulb.
Regards Hans
This subject is a very interesting topic, as it shows the stubbornness of people (myself included, no doubt) in a subject that isn't all that important.
It is similar to theological discussions. Get two parties together to discuss a particular esoteric, unimportant religious subject, and they so vehemently oppose each other they forget to keep in perspective how important the answer is, anyway.
Maybe I can add a little practicality to the subject, if we can all set our emotions aside for a minute. This doesn't have to be as difficult as we are making it.
To start with, I should mention that my preference is for regulated constant current. Further, I would like to remind us all that tube LIFE is not the only issue here. As put forth in one of my prior posts, there are many factors that go into the choice. Too much emphasis is placed on tube LIFE.
So, my take is this:
If you want a simple, time-tested method that works, go with unreg AC voltage straight from the xfmr. One will find thousands of testimonials that this works A-OK.
If you want improved performance from line voltage changes, contact resistance, harmonics, noise, hum, etc., you need some sort of regulation. This should not be much of a debatable issue, other than the discussion of HOW MUCH improvement is actually gained.
I believe filtering ALONE is a waste of $$$ for the capacitance you need. I can provide probably 1/10 the amount of capacitance, follow up with a regulator, and provide equal or better 'filtering', for less or equal $$$. Again, I think this is not very debatable, especially since you still have an unreg supply from line voltage changes.
In the area of constant voltage vs constant current, CC wins. Why? When you weigh ALL (read: ALL, not just tube life) CC provides every benefit of CV, with the additional benefit of inrush limiting. If you take CV and add to it a current limiting feature, my response is to prove the additional complexity of CV with CL has any benefit over straight CC. Adding complexity and parts count is almost never a benefit. CC in this example would be simpler and equivalent.
Finally, the resistance of the filament itself is a moot point, as all the methods above have no control over it. A study of the resistance change of the filament compared to the radiated heat will eventually reach an equilibrium. In other words, no matter what method one uses, eventually the tube will find its happy spot between resistance change and effective heat transfer. All we can change is either V-squared or I-squared. R is not up to us.
It is similar to theological discussions. Get two parties together to discuss a particular esoteric, unimportant religious subject, and they so vehemently oppose each other they forget to keep in perspective how important the answer is, anyway.
Maybe I can add a little practicality to the subject, if we can all set our emotions aside for a minute. This doesn't have to be as difficult as we are making it.
To start with, I should mention that my preference is for regulated constant current. Further, I would like to remind us all that tube LIFE is not the only issue here. As put forth in one of my prior posts, there are many factors that go into the choice. Too much emphasis is placed on tube LIFE.
So, my take is this:
If you want a simple, time-tested method that works, go with unreg AC voltage straight from the xfmr. One will find thousands of testimonials that this works A-OK.
If you want improved performance from line voltage changes, contact resistance, harmonics, noise, hum, etc., you need some sort of regulation. This should not be much of a debatable issue, other than the discussion of HOW MUCH improvement is actually gained.
I believe filtering ALONE is a waste of $$$ for the capacitance you need. I can provide probably 1/10 the amount of capacitance, follow up with a regulator, and provide equal or better 'filtering', for less or equal $$$. Again, I think this is not very debatable, especially since you still have an unreg supply from line voltage changes.
In the area of constant voltage vs constant current, CC wins. Why? When you weigh ALL (read: ALL, not just tube life) CC provides every benefit of CV, with the additional benefit of inrush limiting. If you take CV and add to it a current limiting feature, my response is to prove the additional complexity of CV with CL has any benefit over straight CC. Adding complexity and parts count is almost never a benefit. CC in this example would be simpler and equivalent.
Finally, the resistance of the filament itself is a moot point, as all the methods above have no control over it. A study of the resistance change of the filament compared to the radiated heat will eventually reach an equilibrium. In other words, no matter what method one uses, eventually the tube will find its happy spot between resistance change and effective heat transfer. All we can change is either V-squared or I-squared. R is not up to us.
zigzagflux said:
Ok, please show me your CC, how is it simplier than my CV+CL regulator.
Copying here, for your convenience:
Dependence of resistance on filament temperature impacts on temperature, temperature impacts on emission, emission impacts on everything.
With CC fluctuations of temperature will be AMPLIFIED. In case of CV they will be damped. In case of negative dynamic resistance temperature of cathode, i.e. emission, may be stabilized! I told already about my tool to check calibration of vagon axle - bearing distant temperature sensor.
When I can't live without stabilized tube parameters, I use a CV. That has CL that helps to prolong tube's life. For me it is secondary, primary I don't like constant drift of parameters caused by current stabilizers that were better than nothing when radio receivers were powered from very different sources. If you find any CC tube in some Collins R-390A, it is there probably to preserve historical look and feel since stability of home power is enough for variable oscillator's tube.
.
rdf said:
Yes, but not only, you forgot about voltage that is primary as well. In case of CC increase in temperature increases resistance, it increases voltage that increases temperature, and so on, until it not stop increasing due to high, but finite dynamic resistance.
It works like a positive feedback by temperature.
Quote:
"Ok, please show me your CC, how is it simplier than my CV+CL regulator."
I think that's a little simpler.
You also provided no specifics as to what your 20mA CCS is comprised of. Could be zener-resistor-pnp, or possibly an expensive constant current diode, I don't know.
Fundementally, our heatsinking requirements will be identical, as it comes down to output current * pass headroom.
Both of our schematics did not address the input cap bank or possibly some output filtering (mine uses a PE cap across the output). I simply wanted to compare apples to apples.
Also, I fully understand your argument about positive temperature dependence, my point is it really doesn't matter.
Quote:
"Dependence of resistance on filament temperature impacts on temperature, temperature impacts on emission, emission impacts on everything."
You still cannot control those impacts with CV or CC. If you have four EL34 filaments, and one of them has a higher resistance (or temperature coefficient of resistance) than the others, that tube will get less (more with CC) heat than the others, and you cannot change this without individual control of each heater.
A heater will increase R with temperature, but a CCS does NOT promote a thermal runaway condition, as you claim. There is a limit whereby the equilibrium between the applied and emitted energies will balance, and the heater will stabilize. This is what actually happens, not thermal runaway.
So both CV and CC have a self-regulating mechanism, but they do it differently. CV regulates itself by coefficient of resistance, CC does it by heat gradient (for lack of a better term).
"Ok, please show me your CC, how is it simplier than my CV+CL regulator."
I think that's a little simpler.
You also provided no specifics as to what your 20mA CCS is comprised of. Could be zener-resistor-pnp, or possibly an expensive constant current diode, I don't know.
Fundementally, our heatsinking requirements will be identical, as it comes down to output current * pass headroom.
Both of our schematics did not address the input cap bank or possibly some output filtering (mine uses a PE cap across the output). I simply wanted to compare apples to apples.
Also, I fully understand your argument about positive temperature dependence, my point is it really doesn't matter.
Quote:
"Dependence of resistance on filament temperature impacts on temperature, temperature impacts on emission, emission impacts on everything."
You still cannot control those impacts with CV or CC. If you have four EL34 filaments, and one of them has a higher resistance (or temperature coefficient of resistance) than the others, that tube will get less (more with CC) heat than the others, and you cannot change this without individual control of each heater.
A heater will increase R with temperature, but a CCS does NOT promote a thermal runaway condition, as you claim. There is a limit whereby the equilibrium between the applied and emitted energies will balance, and the heater will stabilize. This is what actually happens, not thermal runaway.
So both CV and CC have a self-regulating mechanism, but they do it differently. CV regulates itself by coefficient of resistance, CC does it by heat gradient (for lack of a better term).
TECHNIQUES TO EXTEND THE SERVICE LIFE OF HIGH POWER VACUUM TUBES
This is largely about direct heated tubes, though the heater itself is of similar construction in oxide cathode tubes. In a direct heated tube, inrush current stresses the filament in proportion to the cube of the current. Constant current is a definite plus for your spendy DHTs.
zigzagflux said:
It is not a runaway, it is a positive feedback that increases gradient of temperature if someting caused it to change (for example, heat from power transformers).
Eureca! We invented a heat amplifier, let's patent it!
Possible amplification.
A tube with 2 filaments close to each other. First one connect to the antenna socket, in series with it.
Second one connect to CCS and voltmeter.
The voltmeter will show RF current in antenna.
astouffer said:
So, you are against limiting of inrush current? Come on, it is cheap to do...
yagoolar said:
I've posted a diagram of the conclusion already... Stable voltage means stable power on heater and stable parameters of amplifier. Also, voltage regulation prevent overheating that shortens life in terms of good emission. Limited inrush current prolonges life of filaments and lowers possibility of shorting filament to cathode.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.