I forgot to mention, that during startup, due to the huge temperature gradients, unnecessary stresses occur inside the valve.
This can be seen and sometimes even listen with certain valves.
Solution : Soft start in heaters PSU.
Protecting a single valve, single phenomenon: 15 cents
Keep all valves happy: A good power supply 😀
Attachments
Well, I wrote it correctly, I think it it is understood.
I forgot the infrared photoemission. 😱
The next time that take the guitar*, I will take this into account, thanks !
* Folk musical improvisation used in these latitudes. 😀
Nothing to be thankful, we are here to help each other. 🙂
Ufil = [(R4+R5)/R5)] Ureff
Ureff = 1.2V (LM385 1.2)
Ureff = 2.5V (LM385 2.5)
Hint: If a 78XX seems too much, you can use a zener.
Any low noise, single supply, op amp can work.
Last edited:
Not to be annoying, just out of curiosity. 🙂
I'm out of the system, I didn't even credit card I have.
Someone has Physical Review subscription?
http://www.google.com.ar/url?sa=t&r...67V7vDy67O59o9uVe-1thaw&bvm=bv.42261806,d.eWU
http://www.google.com.ar/url?sa=t&r...gtYJ5BgGAgztSjJeDUW1yiQ&bvm=bv.42261806,d.eWU
Thanks in advance.
I'm out of the system, I didn't even credit card I have.
Someone has Physical Review subscription?
http://www.google.com.ar/url?sa=t&r...67V7vDy67O59o9uVe-1thaw&bvm=bv.42261806,d.eWU
http://www.google.com.ar/url?sa=t&r...gtYJ5BgGAgztSjJeDUW1yiQ&bvm=bv.42261806,d.eWU
Thanks in advance.
This is unrelated- it's for high pressure systems like in CVD or plasma or ion bombardment, not high vacuum.
Maybe this is not unrelated, and it is only to describe the phenomenon
http://144.206.159.178/FT/68/73201/1254249.pdf
A CRT has indirectly heated cathode, then the extrapolation is valid, at least in first approximation, to an ordinary vacuum tube.
By definition
E = - ∇φ = dφ/dx
Integrating
V = Δφ = ∫dφ = E ∫dx = E d
Then
E = V/d
For a 20-inch CRT, as I see daily
Vak ≈ 25 KV dak ≈ 40 cm Vg1 = 0 V
Then
Eak ≈ 62.5 KV/m 😉
For 12AU7
Vgk ≈ 40 V (DF dixit) dgk ≈ 0.17 mm *
Then
Egk ≈ 235 KV/m 😀
* History of Electron Tubes - Google Libros
For WE416C
Vgk ≈ 40 V dgk ≈ 20 µm (Morgan Jones dixit)
Then
Egk ≈ 2 MV/m
Well, after all I am not so coward. 😎
And if I were the cathode, would continue praying. 😀
Well, energy is another matter, but can not deny that the numbers scare. 😀
http://144.206.159.178/FT/68/73201/1254249.pdf
A CRT has indirectly heated cathode, then the extrapolation is valid, at least in first approximation, to an ordinary vacuum tube.
By definition
E = - ∇φ = dφ/dx
Integrating
V = Δφ = ∫dφ = E ∫dx = E d
Then
E = V/d
For a 20-inch CRT, as I see daily
Vak ≈ 25 KV dak ≈ 40 cm Vg1 = 0 V
Then
Eak ≈ 62.5 KV/m 😉
For 12AU7
Vgk ≈ 40 V (DF dixit) dgk ≈ 0.17 mm *
Then
Egk ≈ 235 KV/m 😀
* History of Electron Tubes - Google Libros
For WE416C
Vgk ≈ 40 V dgk ≈ 20 µm (Morgan Jones dixit)
Then
Egk ≈ 2 MV/m
Well, after all I am not so coward. 😎
And if I were the cathode, would continue praying. 😀
Well, energy is another matter, but can not deny that the numbers scare. 😀
Last edited:
I think you've used a more complex calculation than you need and plugged incorrect numbers into it. 😀
More to the point is the tendency to arc because of the point-like nature of the grid. And that's where a voltage clamp can really help for prevention. If 90V seems too much for you (that's what I use for ImPasse and HMV preamps with not a single failure yet, and there's a few hundred of them out there), use a reverse biased diode to limit it to 0.7V. Far, far cheaper than slow ramping supplies and works just as well- or better.
More to the point is the tendency to arc because of the point-like nature of the grid. And that's where a voltage clamp can really help for prevention. If 90V seems too much for you (that's what I use for ImPasse and HMV preamps with not a single failure yet, and there's a few hundred of them out there), use a reverse biased diode to limit it to 0.7V. Far, far cheaper than slow ramping supplies and works just as well- or better.
well , I sometimes deliberately err in more complicated direction , especially in infamous Babelfish stable members case 
;
but - that's only because I can and choose to , not because I read somewhere on the web that that way is better
there are myriad of ( especially Philips - with more exposed filament from cathode cylinder ) , toobz - totally unprotected from on surge , be it heater or Ub PSU ; they are still good and strong , especially ones where real engineering tricks are used in circuit - proper voltages , proper potentials
so , just to be clear - heater elevating and g-k limiters are good and clever praxis
cathode stripping is in 95% of cases just internet hype ; real example of amateurs to amateurs praxis
;but - that's only because I can and choose to , not because I read somewhere on the web that that way is better
there are myriad of ( especially Philips - with more exposed filament from cathode cylinder ) , toobz - totally unprotected from on surge , be it heater or Ub PSU ; they are still good and strong , especially ones where real engineering tricks are used in circuit - proper voltages , proper potentials
so , just to be clear - heater elevating and g-k limiters are good and clever praxis
cathode stripping is in 95% of cases just internet hype ; real example of amateurs to amateurs praxis
I honestly do not know an easier way to calculate, even more, I thik the right thing would have been to use the Fermi-Dirac distribution.
I use it several times, once to determine the work function of a cathode, and I must confess that it is a mess. 🙄
The numbers were suggested by DF, Vgk = 40 V instead of 90 V or so.
There's a guy out there that broke an ECC82, and found dgk = 1/20 mm, I found a book with dgk = 0.17 mm.
For WE416C, the very Morgan Jones says dgk = 20 µm.
For a 20 inch CRT, dak measured myself with a tape measure, and I've had some close encounters with Vak = 25 KV, ouch !
If you have a better way, or perhaps experimental data, please post them, only then Ion bombardment would be enlightened. 😉
As a design criterion that is valid, good for you, however, place more than one resistor on the grid, scares me.
I spent years making PSUs occupying half of enclosures, never had problems with hum, buzz, audible noise, even more, I never need to use balanced inputs with 66 dB gain mc preamps.
And I do not want to brag, my knowledge is pretty limited, rather than using the standard KISS, I use WKISICW: Why Keep It Simple If Complicated Works ? 😀
cathode stripping is in 95% of cases just internet hype ; real example of amateurs to amateurs praxis
When I was a child, I thought that valve aging is because the cathode run out of electrons.
Now I find that ion bombardment is a better explanation.AFAIK.
As a TV repairman, I repair a CRT every so often, as the Toshiba's paper explains, ion bombardment occurs even with the CRT at normal operation.
It may be that 95% of cases is a myth, then, by Murphy's law, the remaining 5%, does get to mi. 😀
I think his numbers are OK. Interpretation of what they mean is the issue. It is the E field (and local concentration such as by grid wires) which can directly ionise gas molecules. It is the voltage which gives energy to any ions. 40eV is not a lot of energy. I believe (not sure) that it is the momentum which damages the cold cathode, which would vary as sqrt(V).SY said:
Some people seem to get away with no protection. A neon is a big improvement, and can be regarded as being completely out of the circuit during normal operation (apart from a tiny capacitance). A diode is even safer for the grid, but imposes a small (but boostrapped) non-linear capacitance in normal operation.
If you consider the planes of the cathode and grid equipotential, then the V/cm is very straightforward to calculate and very low numbers. The grid-cathode voltage issue is not one of ion bombardment (the grid cross section is too small for that) but rather arcing.
I don't think the grid to cathode spacing of an ECC83 is 20u! 😀
I don't think the grid to cathode spacing of an ECC83 is 20u! 😀
This is correct, I just suggested it on post#31, thanks !
In the worst case, 2MV/m are under vacuum rupture eg from 20 MV/m to 40 MV/m.
However, I do not like a tiny capacitance, much less a non-linear capacitance in normal operation, I prefer many cheap components in the PSU.
I think SY is somewhat stingy, but keep it quiet. 😀
That is precisely the maximum simplification that I did in my calculations, assume the equipotential surfaces as parallel planes.
Otherwise, you would have to solve the Poisson equation in elliptic cylindrical coordinates, I'll pass, thanks.
BTW. If you use V/cm instead of V/m, your results will be two orders of magnitude smaller than mine...
Count me out to play poker with you. 😀
I swear I did not find a damn data for the distance between cathode and anode.
However, in Toshiba's paper fig.7, ion bombardment starts at something like 100eV, not to say wow, but Ion bombarment stops being a ghost.
Hey, that's for WE416C, for the ECC83, I guess it must be similar to the ECC82.
No, neither videogames with you... 😀
Again, ion bombardment is not the issue for cold start grid-to-cathode, it's arcing you need to worry about.
- Status
- Not open for further replies.