Just took a quick look around my stash and guess what?
Some do have g2 flags and some (most ) don't.
6P28, EL38 and some other sweep type have flags on both g1 and g2, but even my biggest sweeper (PL519) does not.
Some do have g2 flags and some (most ) don't.
6P28, EL38 and some other sweep type have flags on both g1 and g2, but even my biggest sweeper (PL519) does not.
If we confine the discussion to designed-for-audio power pentodes and beam power tubes - then the flags are always on Grid-1.
I checked EL34 (including Mullard, EH, Chinese and Svetlana), 6550 Svetlana, KT66 Marconi-Osram, 6L6 Svetlana, 6П3С-Е (=5881 Russian), 6П14П (=EL84 Russian), EL84 JJ.
Some even have Grid-1 support made out of copper.
I checked EL34 (including Mullard, EH, Chinese and Svetlana), 6550 Svetlana, KT66 Marconi-Osram, 6L6 Svetlana, 6П3С-Е (=5881 Russian), 6П14П (=EL84 Russian), EL84 JJ.
Some even have Grid-1 support made out of copper.
Relevant photo,
6GF7 internals. Can't really see it in the photo, but the bigger section has a gold plated grid. Most likely intended to carry quite a bit of grid current (likely in its intended vertical deflection amplifier service). Bodes well for anyone who might like a class AB2 triode amp with them, say (it's a quirky enough tube that it would probably play well for guitar amp service, if not hi-fi in class 2 operation).
Tim
6GF7 internals. Can't really see it in the photo, but the bigger section has a gold plated grid. Most likely intended to carry quite a bit of grid current (likely in its intended vertical deflection amplifier service). Bodes well for anyone who might like a class AB2 triode amp with them, say (it's a quirky enough tube that it would probably play well for guitar amp service, if not hi-fi in class 2 operation).
Tim
Why Gold Grids?
Why Gold Grids?
SY you say work function.
DF96 (Dr. Dave) you say that hot gold emits few electrons.
Work function is the work required to kick an electron loose.
The Handbook of Chemistry and Physics says that the work function of gold is 5.1 and that work function of nickel is just a red curly hair behind at 5.01. There must still be something else going on like fewer positive ions kicked up into the cloud by gold.
Just scratching my head!
DT
Why Gold Grids?
SY you say work function.
DF96 (Dr. Dave) you say that hot gold emits few electrons.
Work function is the work required to kick an electron loose.
The Handbook of Chemistry and Physics says that the work function of gold is 5.1 and that work function of nickel is just a red curly hair behind at 5.01. There must still be something else going on like fewer positive ions kicked up into the cloud by gold.
Just scratching my head!
DT
Work function has absolutely nothing to do with it.
In oxide coated cathode tubes, the cathode is run at 1050 K or near to. At that temperature the oxide coated cathode has an emission approx 100mA per cm^2 of cathode area (Ref: Electron Emission Coating for the Oxide Cathode, Meltzer C D & Widell Eg, 1961, RCA). We might suppose the grid gets as hot as the cathode due to close proximity. Gold and nickel have a work function about the same as tungsten, and at 1050 K the emission is a few nanoamps per cm^2.
According to Contact Potential And Grid Currents, Schrader E R, 1961, receiving tube grids run at less than 800K, so grid emission without migrated cathode material is very low indeed.
Per the Richardson-Dushman equation, electron emission is sharply dependent on temperature. Tungsten does not have a good work function, but becuase it melts at 3680 K it can be run much hotter than just about anything else and emit electronics quite well.
Grids tend to get contaminated with oxide material vaporised off the cathode. This makes the grid emit electrons and thus make the grid go positive - which can be disasterous. The oxide material does not adhere well to gold, so any vaporised cathode material is more likely to go somewhere else where it does less harm.
In oxide coated cathode tubes, the cathode is run at 1050 K or near to. At that temperature the oxide coated cathode has an emission approx 100mA per cm^2 of cathode area (Ref: Electron Emission Coating for the Oxide Cathode, Meltzer C D & Widell Eg, 1961, RCA). We might suppose the grid gets as hot as the cathode due to close proximity. Gold and nickel have a work function about the same as tungsten, and at 1050 K the emission is a few nanoamps per cm^2.
According to Contact Potential And Grid Currents, Schrader E R, 1961, receiving tube grids run at less than 800K, so grid emission without migrated cathode material is very low indeed.
Per the Richardson-Dushman equation, electron emission is sharply dependent on temperature. Tungsten does not have a good work function, but becuase it melts at 3680 K it can be run much hotter than just about anything else and emit electronics quite well.
Grids tend to get contaminated with oxide material vaporised off the cathode. This makes the grid emit electrons and thus make the grid go positive - which can be disasterous. The oxide material does not adhere well to gold, so any vaporised cathode material is more likely to go somewhere else where it does less harm.
So gold acts indirectly by keeping the grid clean? Sounds plausible.
This may be one of those annoying moments where all the books say "Gold grids emit less" but none of the books say why.
This may be one of those annoying moments where all the books say "Gold grids emit less" but none of the books say why.
No, it's a work function issue. See: Richardson-Dushman equation; the work function is in the numerator of the exponential term.
Platinum is even better than gold in this respect. Nickel is far too reactive.
edit: Some nickel alloys were used for grids, but for reasons other than emission. Nickel was used, though, for the supports.
Platinum is even better than gold in this respect. Nickel is far too reactive.
edit: Some nickel alloys were used for grids, but for reasons other than emission. Nickel was used, though, for the supports.
Plug the numbers into the R-D equation. Gold plating molybdenum (the most common grid material) drops emission by an order of magnitude, especially if the grids are being run warm. See Kohl, "Materials Technology for Electron Tubes," p200 and onward. He specifically cites the gold plating as a way of reducing emission. What is less well appreciated is secondary emission.
Kohl, page 200, and for a considerable amount afterwood, discuuses the properties of copper and copper based alloys. Nothing to do with gold plated grids whatsover.
However, Kohl, on page 247, does say that gold inhibits primary (thermionic) electron emission. Secondary emission in control grids is a non-issue except in large transmitter tubes - these operate the control grid sufficiently positive. On page 241 Kohl explains that control grid electron emission arises from vaporised cathode oxide material landing on the grid.
The work function of a metal is a measure of how hard it is to get that metal to emit electrons. Gold has a high work function, but not significantly diffrent to other metals that can be used in vacuum tubes.
The Richardson-Dushman equation for metals is as follows (emission from non-metals is slighly different):-
I = A . Ao . T^2 . e^[-W/(kB * T)]
where :-
I is emision in amps;
A is area of surface, m^2
Ao is the Richardson-Dushman constant, A/m^2.K^2
T is surface temperature, K;
e is 2.71828...
W is work function of surface, J;
kB is Bloltzman's constant, 1.381x10^-23 J/K
Note the minus sign in front of W. As any schoolboy knows, exp of a negative number gets smaller the more negative the number is.
Sy, if YOU were to put in some numbers into the RD equation, you would quickly discover that at the temperatures in a tube employing oxide coated catheds (1050 K), emission from gold is negligible, and its negiligible for all other metals as well.
Work function has nothing to do with why grids get gold plated in oxide cathode tubes.
In the case of large transmitting tubes, where filaments are thoriated tungsten, temperatures are MUCH higher, and as Kohl points out, a high work function has a slight benefit. But gold work function is not much better than other metals.
Perhaps we have different editions. This is the molybdenum chapter. He explicitly calls out emission as the reason for gold (or platinum) plating on the first page of that chapter.
In any case, because the work function is the denominator of the exponential in the R-D equation, a small difference in WF makes a very large difference in emission. Yes, it's small in any case, which is why we can usually assume ig = 0, but for cases where it can be important (e.g., grids running hot, electrometers, condenser mikes), you can achieve an order of magnitude lower grid emission by plating the molybdenum with a lower work function metal.
Ah, you have the 1951 edition, not the 1960 version. The molybdenum chapter does not seem any more relavent though. The only mention on page 200 about gold is in regard to suppressing secondary emission - relevant to screen grids but not to control grids (large transmitter tubes excepted).
One needs to realise when reading books like Kohl is that it isn't tube type specific. For instance, in the same paragraph he goes on to say that molybdenum is used to make anodes. True, but not for oxide cathode tubes that this thread is about. In such tubes nickel, nickel/steel, or nickel/copper laminates are the go.
The work function of gold 5.1 eV is the same as for nickel, and not much better than that of molybdenum, 4.9 eV. Using T = 1050 K and kB = 8.617x10^05, we see that the exponent varies from about -54 to -56, which changes the emission by one order of magnitude. But how many grids are molybdenum anyway? Molybdenum was used for grids in the early years but has been obsolete for grids in oxide cathode tubes since at least the 1950's.
Nickel, which has the same work function as gold, has ideal mechanical properties to make grids from, and has always been cheap. Nickel cladding would also be cheap. Some grids in power tubes have been made from Hastelloy or Dowmo.
Last edited:
Thermionic emission is described by Richardson-Dushman equation
J0 = A0 T² exp ( - e φ / k T )
Then, thermionic emission can be reduced increasing the work function or lowering the temperature.
As I said before
J0 = A0 T² exp ( - e φ / k T )
Then, thermionic emission can be reduced increasing the work function or lowering the temperature.
As I said before
Unfortunately, Nickel has an electron affinity of about 1.157 eV, almost than a half of gold, about 2.308.
Molybdenum is even worse, with an electron affinity of about 0.747 eV.
On the other hand, a clean gold grid can maintain constant its work function, even more, without contamination from cathode, it cannot act like a black body, which in turn implies less heat absorbed, then the lesser temperature.
If we look at the ratio of emission current for two different materials at a constant temperature, we get exp(-q[deltaW]/kT). Plugging in a deltaW of 0.8eV (using Kohn's numbers) and T of 1000K, the ratio for gold to molybdenum (the usual grid wire) is about 0.0001.
If we want to quibble about the temperature and plug in 800K, the ratio is even smaller.
If we want to quibble about the temperature and plug in 800K, the ratio is even smaller.
Let's put some numbers.
Due to electron affinity
Q(Au)/Q(Ni) ≈ 0.5 (*)
From elemental Thermodynamics
Q = C m ∆T
For gold
C=0.129 J/(gºK), δ=19.3 g/cm³
For nickel
C=0.44 J/(gºK), δ=8.9 g/cm³
Supposing identical grids, they must have the same volume, then
Q(Au)/Q(Ni) ≈ 0.636 [∆T(Au)/∆T(Ni)] (**)
Combining (*) and (**)
∆T(Au) ≈ 0.786 ∆T(Ni)
Taking ambient temperature as 20 ºK, with T(Ni)=1050 ºK
T(Au) ≈ 829.58 ºK
Then with k=8.617 x 10⁻⁵ eV/ºK
J(Au) ≈ 3.43 x 10⁻⁸ J(Ni)
Eight orders of magnitude, that's ignoring thermal conductivity:
For gold 317 W/(ºKm), for nickel 90.7 W/(ºKm)
Then, gold grid should be even cooler... 😀
Due to electron affinity
Q(Au)/Q(Ni) ≈ 0.5 (*)
From elemental Thermodynamics
Q = C m ∆T
For gold
C=0.129 J/(gºK), δ=19.3 g/cm³
For nickel
C=0.44 J/(gºK), δ=8.9 g/cm³
Supposing identical grids, they must have the same volume, then
Q(Au)/Q(Ni) ≈ 0.636 [∆T(Au)/∆T(Ni)] (**)
Combining (*) and (**)
∆T(Au) ≈ 0.786 ∆T(Ni)
Taking ambient temperature as 20 ºK, with T(Ni)=1050 ºK
T(Au) ≈ 829.58 ºK
Then with k=8.617 x 10⁻⁵ eV/ºK
J(Au) ≈ 3.43 x 10⁻⁸ J(Ni)
Eight orders of magnitude, that's ignoring thermal conductivity:
For gold 317 W/(ºKm), for nickel 90.7 W/(ºKm)
Then, gold grid should be even cooler... 😀
OK, for molybdenum vs gold
Due to electron affinity
Q(Au)/Q(Mo) ≈ 0.324 (*)
From elemental Thermodynamics
Q = C m ∆T
For gold
C=0.129 J/(gºK), δ=19.3 g/cm³
For molybdenum
C=0.25 J/(gºK), δ=10.28 g/cm³
Supposing identical grids, they must have the same volume, then
Q(Au)/Q(Mo) ≈ 0.97 [∆T(Au)/∆T(Mo)] (**)
Combining (*) and (**)
∆T(Au) ≈ 0.334 ∆T(Mo)
Taken ambient temperature as 20 ºK, with T(Mo)=1050 ºK
T(Au) ≈ 364.02 ºK
Then with k=8.617 x 10⁻⁵ eV/ºK
J(Au) ≈ 2.94 x 10⁻⁴⁹ J(Mo)
I prefer gold 😀
Due to electron affinity
Q(Au)/Q(Mo) ≈ 0.324 (*)
From elemental Thermodynamics
Q = C m ∆T
For gold
C=0.129 J/(gºK), δ=19.3 g/cm³
For molybdenum
C=0.25 J/(gºK), δ=10.28 g/cm³
Supposing identical grids, they must have the same volume, then
Q(Au)/Q(Mo) ≈ 0.97 [∆T(Au)/∆T(Mo)] (**)
Combining (*) and (**)
∆T(Au) ≈ 0.334 ∆T(Mo)
Taken ambient temperature as 20 ºK, with T(Mo)=1050 ºK
T(Au) ≈ 364.02 ºK
Then with k=8.617 x 10⁻⁵ eV/ºK
J(Au) ≈ 2.94 x 10⁻⁴⁹ J(Mo)
I prefer gold 😀
No, for similar work functions, it weighs more temperature.
In the case of nickel (e φ = 5.01 eV) vs gold (e φ = 5.1 eV), at the same temperature T=1050 ºK
J(Au) ≈ 0.37 J(Ni)
If we take into account electron affinity, the gold grid is much cooler, then
J(Au) ≈ 1.15 x 10⁻⁷ J(Ni)
Sorry, on post#35 I took for nickel e φ = 4.9 eV, but what is an order of magnitude between friends? 😀
Why silver grids?
Why silver grids?
Looking at the GE 5-Star brochure, Page 16 it says “The critical grids of Five-Star tubes are plated with gold and silver. This minimizes emission during life.” http://www.tubebooks.org/tubedata/GE_5star.pdf
Silver has a lower work function!
Silver has greater Thermal Conductivity!
Silver
Electronegativity (Pauling scale):5 1.93
Electropositivity (Pauling scale): 2.07
Electron Affinity:6 1.302 eV
Oxidation States: +1
Work Function:7 4.64 eV = 7.43328E-19 J
Thermal Conductivity: 429 (W/m)/K, 27°C
Gold
Electronegativity (Pauling scale):7 2.4
Electropositivity (Pauling scale): 1.6
Electron Affinity:8 2.30863 eV
Oxidation States: +3,1
Work Function:9 5.32 eV = 8.52264E-19 J
Thermal Conductivity: 317 (W/m)/K, 27°C
Nickel
Electronegativity (Pauling scale):6 1.91
Electropositivity (Pauling scale): 2.09
Electron Affinity:7 1.156 eV
Oxidation States: +2,3
Work Function:8 5.15 eV = 8.2503E-19 J
Thermal Conductivity: 90.7 (W/m)/K, 27°C
The above data comes from this site. SILVER (Ag) • Element47 • Physical Properties/Chemical Properties • Description • Uses/Function • Isotopes • Thermochemistry • Reactions • Electron Configuration and Bonding
Change the last two letters for a different element ie. AG, Au, or Ni
DT
Why silver grids?
Looking at the GE 5-Star brochure, Page 16 it says “The critical grids of Five-Star tubes are plated with gold and silver. This minimizes emission during life.” http://www.tubebooks.org/tubedata/GE_5star.pdf
Silver has a lower work function!
Silver has greater Thermal Conductivity!
Silver
Electronegativity (Pauling scale):5 1.93
Electropositivity (Pauling scale): 2.07
Electron Affinity:6 1.302 eV
Oxidation States: +1
Work Function:7 4.64 eV = 7.43328E-19 J
Thermal Conductivity: 429 (W/m)/K, 27°C
Gold
Electronegativity (Pauling scale):7 2.4
Electropositivity (Pauling scale): 1.6
Electron Affinity:8 2.30863 eV
Oxidation States: +3,1
Work Function:9 5.32 eV = 8.52264E-19 J
Thermal Conductivity: 317 (W/m)/K, 27°C
Nickel
Electronegativity (Pauling scale):6 1.91
Electropositivity (Pauling scale): 2.09
Electron Affinity:7 1.156 eV
Oxidation States: +2,3
Work Function:8 5.15 eV = 8.2503E-19 J
Thermal Conductivity: 90.7 (W/m)/K, 27°C
The above data comes from this site. SILVER (Ag) • Element47 • Physical Properties/Chemical Properties • Description • Uses/Function • Isotopes • Thermochemistry • Reactions • Electron Configuration and Bonding
Change the last two letters for a different element ie. AG, Au, or Ni
DT
- Status
- Not open for further replies.