Yes, any current path differences are easily distributed quickly across the near zero Ohms plate metal, or Carbon plate.
We do not normally saw plates in half, and then measure the different currents in each half. But if we did, we would see the current differences (use lots more than just two halves, try using 10 sections).
Even distribution of electron paths?
Lots of tubes that are used beyond the maximum rated plate dissipation; as well as tubes that go into thermal run-away . . .
They exibit a common phenomena . . . Red Plating.
Typically, the red plating hot spot is in a small area where most of the electrons hit the plate, cause lots of secondary emissions, etc.
Often, the red plate hot spot is on only one side of the plate, but the other identical side of the plate does not have a hot spot. Why?
Well, perhaps the spacings from cathode to grid, grid to screen if present, grid or screen to plate are not equal from one side of the plate to the other side of the plate.
Un-equal electron path lengths.
The voltages between unipotential cathodes, unipotential grids, unipotential screens, and unipotential plates are equal.
But those equal voltages are not the same voltage gradients, when the path lengths are un-equal.
Hot spots can make screens collapse.
Learning about various failures of linearity, hot spots, etc. is helpful so that we can analyze what went wrong, and make things better next time.
Back to Fields and Physics.
We do not normally saw plates in half, and then measure the different currents in each half. But if we did, we would see the current differences (use lots more than just two halves, try using 10 sections).
Even distribution of electron paths?
Lots of tubes that are used beyond the maximum rated plate dissipation; as well as tubes that go into thermal run-away . . .
They exibit a common phenomena . . . Red Plating.
Typically, the red plating hot spot is in a small area where most of the electrons hit the plate, cause lots of secondary emissions, etc.
Often, the red plate hot spot is on only one side of the plate, but the other identical side of the plate does not have a hot spot. Why?
Well, perhaps the spacings from cathode to grid, grid to screen if present, grid or screen to plate are not equal from one side of the plate to the other side of the plate.
Un-equal electron path lengths.
The voltages between unipotential cathodes, unipotential grids, unipotential screens, and unipotential plates are equal.
But those equal voltages are not the same voltage gradients, when the path lengths are un-equal.
Hot spots can make screens collapse.
Learning about various failures of linearity, hot spots, etc. is helpful so that we can analyze what went wrong, and make things better next time.
Back to Fields and Physics.
BTW
I dont know the type of DH cathode with "rounded" shape?
Simply it is not acheivable?
.
Grids ould be in sort of rouded shape like spring for instance.
Anodes canbe formed in rounded barrel too.
.
But for the cathode, to be in circular shape, on some length, it should be:
a) single wire from bottom to top.
b) same as "spring" grid.
.
Only on this way, when all 3 electrodes has rouded shape, it could be an option to talk about rounded shape?
.
With indirectlu heated cathodes, the cathode can be in barrel. as most of IH tubes has?
I dont know the type of DH cathode with "rounded" shape?
Simply it is not acheivable?
.
Grids ould be in sort of rouded shape like spring for instance.
Anodes canbe formed in rounded barrel too.
.
But for the cathode, to be in circular shape, on some length, it should be:
a) single wire from bottom to top.
b) same as "spring" grid.
.
Only on this way, when all 3 electrodes has rouded shape, it could be an option to talk about rounded shape?
.
With indirectlu heated cathodes, the cathode can be in barrel. as most of IH tubes has?
45,
I like your statement:
"Then add the boundary conditions with stray fields to the planar and cylindrical types (especially taking into account how much surface is involved in relation to the ideal one) and see which one looks better."
Some tubes have Very much larger ratios of the desired electron paths, versus un-desired electron paths.
The True Planar Triode 416A is one of those.
I did not trust the drawings, I pulled one apart to see for myself.
The 6DJ8 is sometimes called a dual planar triode. It does not have nearly as good of a ratio of desired electron paths versus un-desired electron paths.
It is Not a true planar triode.
It is a Frame Grid dual triode.
I like your statement:
"Then add the boundary conditions with stray fields to the planar and cylindrical types (especially taking into account how much surface is involved in relation to the ideal one) and see which one looks better."
Some tubes have Very much larger ratios of the desired electron paths, versus un-desired electron paths.
The True Planar Triode 416A is one of those.
I did not trust the drawings, I pulled one apart to see for myself.
The 6DJ8 is sometimes called a dual planar triode. It does not have nearly as good of a ratio of desired electron paths versus un-desired electron paths.
It is Not a true planar triode.
It is a Frame Grid dual triode.
jcalvarez,
I am attaching a drawing.
Look at the cathode, each side is relatively flat.
Look at the curve shape of grid g1
Look at the curved shape of grid g2
Look at the curved shape of the plate (in the paths that are set by the beam formers).
Look at the electron path lengths from the cathode all the way to the plate.
Are the electron path distances equal from cathode to g1, g1 to g2, g2 to the plate?
It seems that reasonable care in design and production have made the majority of the electron paths roughly equal through each of the field potentials along the way (cathode to g1, g1 to g2, g2 to plate).
Does that answer your question?
Equal lengths are good for RF, OK. But we are talking about audio, where the Transit Time does not apply.
But, Linearity Does count. We get good linearity because of the similary paths and similar path lengths of all the electrons.
Linearity, not just for RF, but Linearity for audio.
Does that answer your question(s)?
"Secrets" of dead engineers; Western Electric, RCA, EiMac, etc.
I am attaching a drawing.
Look at the cathode, each side is relatively flat.
Look at the curve shape of grid g1
Look at the curved shape of grid g2
Look at the curved shape of the plate (in the paths that are set by the beam formers).
Look at the electron path lengths from the cathode all the way to the plate.
Are the electron path distances equal from cathode to g1, g1 to g2, g2 to the plate?
It seems that reasonable care in design and production have made the majority of the electron paths roughly equal through each of the field potentials along the way (cathode to g1, g1 to g2, g2 to plate).
Does that answer your question?
Equal lengths are good for RF, OK. But we are talking about audio, where the Transit Time does not apply.
But, Linearity Does count. We get good linearity because of the similary paths and similar path lengths of all the electrons.
Linearity, not just for RF, but Linearity for audio.
Does that answer your question(s)?
"Secrets" of dead engineers; Western Electric, RCA, EiMac, etc.
Yes
It seems like from the same "family" like 100TH, 250TH etc
I ment these types of tubes when mentioning rounded "spring" grids and barrel anodes.
I saw this on these tubes. I remember this...
But I didn't payed attention about cathode?
Some RF tubes with birdcage grid(s) and circular plates, have a very interesting cathode structure:
First, think of cardboard that has a flat outer layer, wavy inner layer, and then the other flat outer layer.
Now . . .
Imagine taking the wavy inner layer, and forming it into a circle (a wavy circle).
Some of the cathode is closer to g1, and some of the cathode is further from g1 (a wavy surface on the circular cathode's face).
I am not sure, but I think the intent was to line up the peaks with the openings between the grid wires of the birdcage grid, g1.
Focused fields?
How about that for special attention to design?
First, think of cardboard that has a flat outer layer, wavy inner layer, and then the other flat outer layer.
Now . . .
Imagine taking the wavy inner layer, and forming it into a circle (a wavy circle).
Some of the cathode is closer to g1, and some of the cathode is further from g1 (a wavy surface on the circular cathode's face).
I am not sure, but I think the intent was to line up the peaks with the openings between the grid wires of the birdcage grid, g1.
Focused fields?
How about that for special attention to design?
Very interesting, but the still unanswered question is: does that improve sound?
By what mechanism?
The answer is not a random unsubstantiated statement as "it sounds good because it is cylindrical" because that is as solid as "it sounds good because it was made in winter" or any similar one, pulled out of the blue.
The real answer is by showing datasheets, curves, measurements
If we want to introduce the "human factor", fine, all we need are some double blind comparisons.
Good luck with any of that 😄😄😄
By what mechanism?
The answer is not a random unsubstantiated statement as "it sounds good because it is cylindrical" because that is as solid as "it sounds good because it was made in winter" or any similar one, pulled out of the blue.
The real answer is by showing datasheets, curves, measurements
If we want to introduce the "human factor", fine, all we need are some double blind comparisons.
Good luck with any of that 😄😄😄
The real answer is that the round plate family is more uniform in characteristics which means that is also more predictable. No funny things happening.
IME, the lower typical low amount of higher harmonics in signal triodes or triode-strapped pentodes with largely dominant second harmonic, the very similar behaviour makes things like 2nd harmonic cancellation possible without significant tradeoffs and resulting in lower THD overall. All this is more complicated or even impossible with other combinations. Still possible with mixed tube types combinations but not really so easy to predict.
And if you get the same distortion, at least at low-to-medium level, of a feedback amplifier without applying any feedback or less distortion than another zero fbk amp it DOES sound better. It's not subjective. Everyone can tell the difference.
Last edited:
I said 6sj7 because of the lack of a top cap. But yes, that was the original order. 6j7 to 6c5 to Gu50 to 6sj7.
If round plates always offer advantages why do so few tubes used in audio use them? Once again, the engineers that designed tubes were no dummies. Surely they would have stuck with round plates for everything if they really did result in better performance if for no other reason than to sell more.
To my knowledge, only NU and Tung Sol made 6sn7 with round plates and then only for a short amount of time. Curiously all the manufacturers of 6sl7 kept with the round plates for quite a bit longer. The 6sn7 is well known as a linear tube and I believe the 6sl7 is equally well regarded. Personally I do not find either the round plate 6sn7 to be any ”better” and I prefer other brands more. Why would companies switch to flat plates for one tube and keep round for another? Makes one think that there might have been an advantage for certain designs but not others.
To my knowledge, only NU and Tung Sol made 6sn7 with round plates and then only for a short amount of time. Curiously all the manufacturers of 6sl7 kept with the round plates for quite a bit longer. The 6sn7 is well known as a linear tube and I believe the 6sl7 is equally well regarded. Personally I do not find either the round plate 6sn7 to be any ”better” and I prefer other brands more. Why would companies switch to flat plates for one tube and keep round for another? Makes one think that there might have been an advantage for certain designs but not others.
