Yes it is. Fig 5.2.5 is also clearly labelled as potential distribution in a planar diode - ie positive or accelerating external field near the cathode. In this scenario there is a potential minimum due to the interaction between barrier potential at cathode surface and the external accelerating field. This effect is described very well in Harmmann/Waganer Vol2 1.2, pages 19 et seq equations (31) - (35) and has nothing to do with an 'electron cloud'. And what we're mostly interested in is behaviour in a retarding field in grid valves in normal operation in any event.
The term 'space charge' means all things to all men, I'm sure that's a big part of the problem here. Space charge properly used should always be read as 'charge carriers' - as such electrons can't ever 'leave' space charge, they comprise it. Anode current is conveyed by space charge, and space charge smoothing is a phenomenum applicable to anode current generally, since carriers interact in all space charge here. Any current in a valve is conveyed by space charge, subject to space charge smoothing. Space charge isn't just a bunch of free electrons hanging about near the cathode, though that would also comprise space charge since it would be made up of carriers. I really think straightening out what is meant by space charge in literature would really help a common understanding.
If there were a cloud which varied in density, it would also vary in charge hence would vary in potential hence would vary mu........ its location would also shift with grid potential, BTW........but just roughly calculate likely potential associated with it, that should convince you against I think....DF96 said:
See my reply to Keit above, there's a dip for diodes and positive grid accelerating field conditions but only a cousin of charge carrier generation. See Dow (1937) Fig 4(d) Ch5 etc for various triode conditions, and Harmann/Waganer 1951 Vol2 pages 19 et seq for explanation.DF96 said:
Thanks, Merlinb. It's good. Discussions of potential diagrams seem to confirm other texts, but with some good detail. I'm still not seeing what one might expect if there were a thumping cloud of charge near the cathode.......
In this sentence from section 5.2, I substituted 'charge carriers' for 'space charge'. I think this helps reveal proper sense when reading vintage texts:
Grid controlled tubes are almost always operated with the current drawn from the cathode charge-carrier limited since only then is it possible for the grid to act effectively as a control electrode.
Stated this way, much also becomes suddenly familiar to the semiconductor world I think. Misunderstanding space charge is probably responsible for much modern day confusion about 'electron clouds' IMO.
Last edited:
Thanks, that text is excellent. Fig 4 as you say is for diodes and shows expected characteristic potential dip for accelerating fields as per my last post to you with same explanation see H/W 1951 Vol2 1.2 pages 19 et seq. But more interestingly it also shows no such dip for retarding fields, Fig 4 curve C, ie as per grid controlled valves with negative bias !!!
In fairness it will take me a time to read properly and inwardly digest the full text, but a flick through didn't throw any surprises or contradictions.......
That's the crux question...........seems for all practical purposes it might well be, and control exerted by grid potential over carrier release from the cathode, via external field at cathode surface.Keit said:
Re Bigun's post #123:
If an RCA engineer was to say in writing that something is so, I'd put a LOT more faith in that than some turkey in an internet forum - no disrespect to DF96 intended.
1. Those RCA training book/papers were written by senior RCA reseachers at the top of their game. All of them famous names in the industry. Men who had spent many years of their professional life in tube engineering, research, and design.
2. Those RCA training books/papers were written for engineers freshly hired by RCA, and most would have been just out of university, many of them with advanced degrees. You ever worked with such people? I have. And I was once a newly minted honours graduate myself (though I never worked for RCA). Such people are naturally intensely inclined to show how smart they are by pointing out the slightest error or unsubstantiated fact in a lecturer's presentation. Even back in the 1950's and 60's when there was more respect for one's betters than there is now.
Theose RCA traning books were mostly orginally the notes from lectures given in a class room session.
But I take your point that the topic in question in this thread is just a statement in a couple of paragraphs, unaccompanied by any math.
But if the electron cloud doesn't protect the cathode, then what does? The fact that the cathode doesn't last long if the anode voltage is over a certain limit (usually about 500V) is a practical fact. The fact that oxide cathodes don't last long if the tube is operated temperature limitted (saturated) is a practical fact.
So what IS the explanation? Certainly not luckythedog's incoherent nonsense about there being NO electron cloud under normal (unsaturated) conditions. Nothingness can't protect anything.
If an RCA engineer was to say in writing that something is so, I'd put a LOT more faith in that than some turkey in an internet forum - no disrespect to DF96 intended.
1. Those RCA training book/papers were written by senior RCA reseachers at the top of their game. All of them famous names in the industry. Men who had spent many years of their professional life in tube engineering, research, and design.
2. Those RCA training books/papers were written for engineers freshly hired by RCA, and most would have been just out of university, many of them with advanced degrees. You ever worked with such people? I have. And I was once a newly minted honours graduate myself (though I never worked for RCA). Such people are naturally intensely inclined to show how smart they are by pointing out the slightest error or unsubstantiated fact in a lecturer's presentation. Even back in the 1950's and 60's when there was more respect for one's betters than there is now.
Theose RCA traning books were mostly orginally the notes from lectures given in a class room session.
But I take your point that the topic in question in this thread is just a statement in a couple of paragraphs, unaccompanied by any math.
But if the electron cloud doesn't protect the cathode, then what does? The fact that the cathode doesn't last long if the anode voltage is over a certain limit (usually about 500V) is a practical fact. The fact that oxide cathodes don't last long if the tube is operated temperature limitted (saturated) is a practical fact.
So what IS the explanation? Certainly not luckythedog's incoherent nonsense about there being NO electron cloud under normal (unsaturated) conditions. Nothingness can't protect anything.
Last edited:
That is stating the bleeding obvious. Of course a grid swing can't much increase anode current, if the cathode has no more to give. It neither proves nor disproves the existence of and electron cloud, space charge, or collection of charge carriers - whatever you want to call it -it means the same thing.
But many other aspects of tube operation follow from the existence of an electron cloud - I listed the important ones in my previous posts. You still haven't provided a coherent alternative explanation for any of them, let alone all of them. It's no good saying you satisfied yourself. That's like when you claimed to have done a calculation, but are unable to post it.
Member
Joined 2009
Paid Member
Perhaps the electron cloud does offer protection - but it isn't intuitively obvious to me that it can. I haven't seen an analysis or measurements of the energy loss suffered by energetic ions (in the range 100eV to 500eV) with low energy electrons (i.e. the cathode cloud). What are the relevant mechanisms for energy loss from the ion ? - inelastic collisions is a bit of stretch given the enormous mass difference involved in ion-electron collisions. What other energy transfer options do we know about? Is there a mechanism that offers a plausible explanation for the loss of 100's of eV by energetic ions over a relatively short distance ?
Of course, the alternative is that the electron cloud offers negligible protection. And that there is no real protection available to the cathode from any other mechanism. Therefore, the cathode will, over time, become damaged through ion sputtering at a rate that depends on the quality of the vacuum.
Last edited:
The crux of the matter here was identified, by DF96 as I recall, quite early in this thread: Just what is, numerically, the average number or collisions ecountered by an ion, having been accelerated by the accelerating field of the anode? And what fraction of collisions does what? I think I know a general approach to take to calculate these numbers, but I'm too lazy to do it. I haven't seen it done in any book - that's not to say no book or paper has it of course.
Meanwhile, Luckythedog has been trying to claim that no electron cloud exists - a side track.
That the cathode IS protected during normal operation there is no doubt.
It's what a multitude of tube books say, including the RCA book I cited in previous posts. Many books have microphotographs of what happens to oxide cathodes when you raise the anode above about 500 V - a level that imparts enough kinetic energy to the ions to overcome the protection. It is well known that in transmitting tubes where the anode voltage has to be above 500V, they don't use oxide cathodes for that reason - they use the much less efficient thoriated tungsten instead.
Yet, well below 500V the same damage occurs to ixide cathodes when the cathode is not hot enough to produce emission greater than the anode current.
Back to the matter of how an electron cloud could deal with ions: If it is a negative ion (yes, they do exist, an atom or a molecule with the right orbital configuration - eg a metal - can end up with one too many electrons) and these ARE an issue in vacuum tubes - no problem, the dense electron cloud can repel it. It it is a postive ion (lost an electron), it is more complicated. On a fraction of collisions with electrons, it will be charge nuetralised. This is thus not an elastic collision. On a fraction of subsequent collisions, it many get an electron knocked out again - again not an elastic collision. When collisions are not elastic, some of the kinetic energy gets converted to bond energy and vice versa. When collisions occur, the particles fly off again in different directions depending on the collision (as in balls colliding in a glancing blow).
So, for any given atom or ion penetrating the electron cloud between the cathode and grid, we can expect a number of collisions, some elastic (where the light mass of an electron can't do much) and some not elastic (where the electric potential of an electron can make a greater difference), until what begun as a fast moving ion heading straight for the cathode ends up charge nuetralised and slowly moving in some random direction, without enough energy to be re-ionised once more.
Last edited:
Perhaps one of the physicists here would like to explain why luckythedog's expectation of the potential gradient is wrong, in brief terms? Maybe then he'll let go of this bone!
My enquiry is whether a dense free electron cloud near the cathode exists in a normal grid controlled valve in normal operation. As common wisdom, including that asserted by Keit, apparently holds to be true - and I can't find a basis to say that view stands reasonable scrutiny. Quite the opposite.
Obviously, if there is no such discrete cloud or it is of insignificant density/shape, then the only electrons affording protection to the cathode from ion bombardment are in the charge carriers associated with conduction..........
Why ? In the little which seems to be discussed on the topic in serious papers, ion bombardment is acknowledged as an operational phenomenum, but mitigation of it by collisions with charge carriers isn't AFAIK.Keit said:
What might have happened in history is that ion bombardment was sometimes a problem after long term storage, and recommendation emerged to re-activate getter material (inc cathode) before applying B+ . That is 'delay B+ to allow getter material to heat and remove gas atoms from the vacuum, thus helping to avoid formation of ions'. That seems to make more sense IMO. My point with the OP is that it seems unlikely to work very well.......
Seems that reasoning is supposition, or at best 'lore' or common wisdom I think. It isn't discussed in the Harmman/Waganer excellent reference on oxide cathodes Harmann/Waganer 1951 Vol 2, and surely it would be if significant to life/performance of oxide cathodes.Keit said:
OK so this 'cloud' allegedly has associated with it a potential and field strong enough to repel a heavy 300eV ion, but doesn't show up on potential maps and doesn't even feature in discussion of valve operation ?? That's really not credible, IMO.Keit said:
Last edited:
Yes, a reasonable explanation of why a significant lump of free charge apparently doesn't show up in potential gradient diagrams, and for that matter wouldn't be disruptive to operation, should do it. I trust vintage papers to have got it right. Potential gradient diagrams are clear. I take from them, and associated discussion, that charge carrier effects modify potential gradients, of course, but there's no sign of a discrete dense cloud of charge carriers in normal operation of grid controlled valves. I would also accept any serious paper which describes and quantifies any such effect.
But, the force on an electron, and on an ion (which has lost one electron), in the same electric field would be equal (and opposite). Since the electron is so much lighter it would have much greater acceleration. Presumably, given some assumptions about mean free path between collisions, the electrons are therefore travelling much faster than the ions. Momentum is mass x velocity.
Edit: OK I see a flaw here! The electrons in 'the cloud' are not accelerated by the field so much as shot out of the surface by thermionic emission. Whereas the ions have been accelerated by the electric field between anode and cathode (ignoring the grid for now).
Last edited:
But you have been unable to coherently/creditably explain why. You just keep asserting that there isn't a cloud, and keep asserting textbooks support you, which they don't. Just because you can't or won't understand it, it doesn't mean others are wrong.
Exactly. No relatively dense cloud, no effective protection. The cloud is the only difference between space limitted and temperature limiited operation.
That's why noise diodes like the A2087, which are designed for applications in which temperature limitted operation is mandatory, cannot use oxide coated cathodes. They are forced to use the dramatically less efficient pure tungste cathodes, which stand up to it, although their service life is still limitted.
Yes it is. Fo example, I quote from Tomer, 1960, pages 18,19:
"An indirectly heated cathode type tube does not draw its plate current directly from its cathode, but from a revervoir of electrons in the space around the cathode. This resevoir (space charge) is supposed to stay ahaed of the plate current demands. ....
While the cathode is warming up, .... if plate current demands are made on a tube that is temperature limitted, the resevoir of electrons will be completely swept away, leaving the cathode exposed to bomardment by the heavy negative ions that are always present. These negative ions are repulsed by the space charge when it exists, but when it has been drawn away, there is nothing to stop the ions from plunging violently into the cathode coating."
Many other texts say the same thing.
And it's meaningless for you to say that authors were misled by common wisdom or some such, unless you can a) show exactly why they are wrong, and b) provide alternative explanations for all that the electron cloud implies, not just cathode protection from ions.
Gas is and was often a problem sometimes with tubes taken out of long term storage, at least with tubes made before glass/metal sealing with dumet leads was perfected. But that is an entirely separate issue to the well known issue that oxide cathodes are rapidly damaged by operation in the temperture limitted condition - whether they are old, taken out of long storage, or brand new.
That's why noise diodes employ the inefficient pure metal cathodes. It's why transmitting tubes are made with thoriated tungsten cathodes, as the space charge, while good enough to protect oxide cathodes with low anode voltages, can't cope with high energy ions you get with kilovolt anode voltages.
Herrmann/Wagener discusses it in some detail in Vol 1 Sect 12.5 (page 112).
But it does show up in texts. I've given you references. You don't see what you don't want to see, it seems.
Last edited:
I absolutely say no such thing. I say the serious vintage papers have it sorted out, and are right. If anything, it's our interpretation of them that is at issue, as I see it. Especially as to the concept of space charge for example, which based on posts in this forum seems no two posters share a common understanding........
You've confused space charge with 'electron cloud', as previously pointed out.Keit said:
Last edited:
I thought the meaning of space charge was fairly clear: an unusually dense collection of electrons in space (i.e. not on a conductor surface). In a working valve there is usually a space charge near the cathode. In many tetrodes and pentodes there is another (weaker) space charge somewhere near the anode - it is this which suppresses secondary emission from the anode. The main electron stream from the cathode to the anode is not really a space charge as it is too sparse.luckthedog said:
No. Mu depends mainly on geometry, not potential. It is only via 'island effect' (which effectively removes part of the valve from operation) that mu varies with potential. The space charge is effectively a conductor, and it is the ratio of distances between that and the grid and anode (as Keit keeps reminding you) which sets mu.
Who said that the ion is 'repelled'? On the contrary, it is attracted (while it remains an ion) but is then slowed down by many collisions (probably mostly elastic). My recollection is that it does show up on potential maps, and will feature in any detailed discussion of valve operation. Of course, some books might gloss over details so absence of evidence should not be taken as evidence of absence.
Sorry, don't get this at all. Space itself can never be charged. The only way that there can be a 'charge distributed in space' (i.e. a 'space charge') is if there are charges (i.e. charged particles) distributed in that space. (I seem to be getting tautological here!)
Hypothetical bit: The converse is not true. If there were an equal number of + and - charged particles distributed randomly in a space, then the net space charge would be zero. (I know it couldn't stay like this for long - its just hypothetical.)
That's what some people take it to mean.......but it's not 'correct' in the sense of serious vintage authors and in the classical physics sense. Space charge is a concept describing charge carriers generally. All current conveyed in a valve is by charge carriers know as space charge. When 'space charge limited' operation is discussed, what is meant is 'carrier limited' or 'carrier starved' operation, for example. Which might also have some semiconductor meaning as to the concept, for some of us. Anywhere where there are free electrons in a vacuum, there is space charge. It has its own potential gradient depending on density and shape, of course.
The concept at issue is 'virtual' location and 'potential' of the cathode which is a vintage concept to explain adjustment to theoretical mu. My point is that if this were physically associated with an emission based charge cloud (of which there is no vintage mention AFAIK), emission should alter potential gradient ie 'virtual' location and/or potential of cathode and hence mu. But it doesn't (well not much).DF96 said:
Keit did, talking about negatively charged ions. If a dense electron cloud shows up on potential maps, bring it on and let's settle it, as per Merlinb's post - I don't see it.DF96 said:
Yup space charge is a physics concept which simply helps work out potential gradient and interaction with fields in 3D space by treating charge carriers as homogenous and distributed. It's a concept describing charge carriers, in this case when there are no electrons in a volume, space charge is zero. There can be no conduction in valves without it, anode current is conveyed by it. This is the 'correct' meaning of space charge in the context of serious vintage texts.
The hypothetical bit: equal numbers of + and - free charge carriers in space wouldn't hang about long, they'd recombine as you say. But whilst they existed closely in pairs they'd have zero net space charge, except very locally. Check out carrier transport in semiconductors, where the cousins of such phenomena are free charge carriers.
Last edited:
Thanks for clarifying that. I'm sorry if I keep chipping in at an over-simplified level here. But, I think I can now see what you guys are arguing about:
You are saying that thermionic electrons could be prevented from leaving the surface of the cathode by the local electric field (between the anode, and/or grid and the cathode). (Somewhat analogous to the way water boiling is prevent by an increase in air pressure.)
Could we check this by (some clever physicist) calculating the average velocity of a thermionic electron leaving the cathode and seeing how much force would need to be produced by the local field to stop it in its tracks!
(Easy for me to say - I know!)
Last edited:
Member
Joined 2009
Paid Member
I would share the understanding that space charge is defined as net charge per unit volume. There is a net negative space charge inside an operating triode. Here's how I see it
There is a flux of electrons received at the anode. Flux is defined as the number of particles/charges passing through a surface per second. A flux of positive ions would be a current, a flux of negative charges would be a current too, but by convention we define the direction of flow of current to be that of positive charges - and opposite direction to the flow of negative charges.
In physics we say that Flux is conserved meaning that the electrons are neither created or destroyed during their journey from cathode to plate. This is the same as saying that the current that leaves the cathode must all arrive at the anode - current flows in a continuous circuit.
The current flow through the tube is constant across the vacuum, i.e. the flux of electrons is constant across the vacuum. At the anode they are travelling very fast, having been accelerated by the electric field. Near the cathode they are travelling slowly. For the flux to be the same at both electrodes the local density of electrons must be much higher where they are travelling slowly because the same number of electrons per second must leave the cathode as arrives at the anode. This local density of electrons is Space charge. The space charge is highest near the cathode where the electrons travel slowly, and lowest at the anode where they travel quickly.
Electrons inside the cathode have a distribution of energies - a thermal phenomena. Those few electrons with enough energy to escape the collective positive charge of the nuclei of the metal atoms at the surface of the cathode will move into the vacuum. The hotter the cathode the more electrons there are with enough energy (this energy can also be supplied by photons - e.g. ultra-violet light). Once the electrons have left the surface of the hot cathode there is an imbalance of charge between the cathode and the space above it. This acts to attract the electrons back to the metal. There is a sea of boiling electrons at the surface of the cathode and a 'steam' or cloud of electrons above it. This is a well established phenomena.
Because there is a distribution of electron energies inside the metal, there are in principle always going to be some that can escape from the surface. Local electric fields modify the potential barrier that affects their escape but there is always some that escape. There are other mechanisms that allow electrons to overcome the surface potential barrier too - tunnelling effects - which would only be an unnecessary complication to this discussion.
fyi - in terms of mass, a Nitrogen atom hitting an electron is not far off equivalent to a high speed golf ball hitting a mosquito...
There is a flux of electrons received at the anode. Flux is defined as the number of particles/charges passing through a surface per second. A flux of positive ions would be a current, a flux of negative charges would be a current too, but by convention we define the direction of flow of current to be that of positive charges - and opposite direction to the flow of negative charges.
In physics we say that Flux is conserved meaning that the electrons are neither created or destroyed during their journey from cathode to plate. This is the same as saying that the current that leaves the cathode must all arrive at the anode - current flows in a continuous circuit.
The current flow through the tube is constant across the vacuum, i.e. the flux of electrons is constant across the vacuum. At the anode they are travelling very fast, having been accelerated by the electric field. Near the cathode they are travelling slowly. For the flux to be the same at both electrodes the local density of electrons must be much higher where they are travelling slowly because the same number of electrons per second must leave the cathode as arrives at the anode. This local density of electrons is Space charge. The space charge is highest near the cathode where the electrons travel slowly, and lowest at the anode where they travel quickly.
Electrons inside the cathode have a distribution of energies - a thermal phenomena. Those few electrons with enough energy to escape the collective positive charge of the nuclei of the metal atoms at the surface of the cathode will move into the vacuum. The hotter the cathode the more electrons there are with enough energy (this energy can also be supplied by photons - e.g. ultra-violet light). Once the electrons have left the surface of the hot cathode there is an imbalance of charge between the cathode and the space above it. This acts to attract the electrons back to the metal. There is a sea of boiling electrons at the surface of the cathode and a 'steam' or cloud of electrons above it. This is a well established phenomena.
Because there is a distribution of electron energies inside the metal, there are in principle always going to be some that can escape from the surface. Local electric fields modify the potential barrier that affects their escape but there is always some that escape. There are other mechanisms that allow electrons to overcome the surface potential barrier too - tunnelling effects - which would only be an unnecessary complication to this discussion.
fyi - in terms of mass, a Nitrogen atom hitting an electron is not far off equivalent to a high speed golf ball hitting a mosquito...
Last edited:
- Status
- Not open for further replies.