I think, I'm not the first person doing this.
with a simple induction heater, the getter can be re-activated to improve the vacuum.
I have tried this to some very gassy tubes, and once there still a little bit mirror left, in most case, the vacuum can be restored.
it also good to heat the getter for a short moment (about 3 seconds), for those NOS tubes before first time heat up the cathode.
before getter re-heat, a lot holes on mirror.
during the process, the getter should not go too hot (no more than orange - yellow), otherwise it melt
after 5 second heating, the getter mirror become much more thick.
The glowing during (normal circuit) operation is gone.
with a simple induction heater, the getter can be re-activated to improve the vacuum.
I have tried this to some very gassy tubes, and once there still a little bit mirror left, in most case, the vacuum can be restored.
it also good to heat the getter for a short moment (about 3 seconds), for those NOS tubes before first time heat up the cathode.
before getter re-heat, a lot holes on mirror.
during the process, the getter should not go too hot (no more than orange - yellow), otherwise it melt
after 5 second heating, the getter mirror become much more thick.
The glowing during (normal circuit) operation is gone.
Last edited by a moderator:
Hi guys
I use this circuit:
(you do not need to by one, forced air cooling will be sufficient)
the critical thing to make this thing work is:
1. choose the right capacitors. I use 4x 22nF 50V NP0 ceramic capacitor in parallel. I have tried various "plastic" film capacitors (MKP etc.), they got very hot and bulged after 5 minutes. any class 2 ceramic capacitors (X7R etc.) also not work well, I have tried 220nF 250V X7R capacitor, again, very very hot, and they cracked after several time use.
2. coil, use some large size wire.
3. increased voltage --> 25V on L2. (standby current is about 0.8A, loaded current is about 2A)
4. find a good fan (hair dryer in cool air) for it.
I use this circuit:
(you do not need to by one, forced air cooling will be sufficient)
An externally hosted image should be here but it was not working when we last tested it.
the critical thing to make this thing work is:
1. choose the right capacitors. I use 4x 22nF 50V NP0 ceramic capacitor in parallel. I have tried various "plastic" film capacitors (MKP etc.), they got very hot and bulged after 5 minutes. any class 2 ceramic capacitors (X7R etc.) also not work well, I have tried 220nF 250V X7R capacitor, again, very very hot, and they cracked after several time use.
2. coil, use some large size wire.
3. increased voltage --> 25V on L2. (standby current is about 0.8A, loaded current is about 2A)
4. find a good fan (hair dryer in cool air) for it.
i thought you will heat the getter on glass
ZVS Model Low Voltage Induction Heating Machine High-frequency Heating Machine | eBay
ZVS Model Low Voltage Induction Heating Machine High-frequency Heating Machine | eBay
First, thank you for posting that simple example of an induction heater.
Here is a question about the effects upon otherwise good tubes.
Over time the heat bakes adsorbed and absorbed gasses out of the metal elements, the mica, and can break up the alkali oxides on the cathode. These ions then cause grid current, among other problems.
Have you tried baking any obviously non-gassy tubes to see if the performance characteristics change when these ions are bound up in the getter material?
A decent tube tester should be able to measure grid current and that is usually an accurate indicator of how many ions are present in the tube.
Here is a question about the effects upon otherwise good tubes.
Over time the heat bakes adsorbed and absorbed gasses out of the metal elements, the mica, and can break up the alkali oxides on the cathode. These ions then cause grid current, among other problems.
Have you tried baking any obviously non-gassy tubes to see if the performance characteristics change when these ions are bound up in the getter material?
A decent tube tester should be able to measure grid current and that is usually an accurate indicator of how many ions are present in the tube.
thank you,
the induction heating is only targeted for getter. the plate and grid are not effectively heated to a temperature can release gas. and, mica is not metal.
for the new tubes the grid current can only be generated by the gas ionization in the tube. for heavily used ones, particular high Gm tubes (first grid is very close to cathode and heated by cathode), the first grid had been contaminated by the vaporized cathode coating, make it become "cathode" too, which, obviously results negative grid current.
Yes, for those non-gassy tubes, no change on emission or grid current after getter activation.
note, the induction heating is affect to the getter ring, not the "mirror", because the "mirror" is a layer of amorphous metal, the conductivity is relatively bad.
it's difficult to say.
with a current / voltage meter connect to a operating tube between first grid and cathode. for those good tubes, the first grid current is positive ( current flow into grid, negative voltage potential to cathode). gassy tubes, the first grid current is negative ( current flow out grid, positive voltage potential to cathode), this is also the reason of theme-runoff happen to some gassy tubes.
make a comparison between known good tube and test subject is the better way to determine how good / bad it is.
Yes, of course, I do understand this. I did not clearly explain the basis my question.
The normal operation of a tube heats certain structures, and these emit gasses absorbed or adsorbed during manufacturing. My point was that even the best tubes have some residual gas that is baked out by normal operation, not liberated by by rejuvenating the getter. Much work went into creating tube materials with lower contaminants, particularly gasses, because these were emitted over time and thus could not be easily removed either during the inital pumping down or getter flashing, as they were not yet at the surface.
So my question is: what does rejuvenating do to the minimal levels, but still important, liberated by normal operation of the tube when it is heated as part of its normal operation? As well as any gasses baked out of the cathode by destruction of the alkali oxides, of course.
The ions are not restricted to vaporized and ablated cathode material (although this is a source of liberated oxygen) or heater material. The other materials in the tube contain contaminants, such as gasses, which are liberated in normal operation as they migrate to the surface.
No change at all? That's interesting. Are you able to measure small changes in currents? The change may be too subtle to be visible on a tube tester.
Any of the getter material should be capable of vaporizing with inductive heating, it need not be attached to the ring. At least as I understand how the getter functions.
The getter ring itself is merely a support structure. Umbreit's original getter material was barium-aluminum, to ensure it reacted with everything. Getter material was applied as a paste to the inside of the envelope or put on the support ring, but the metal itself is inductively heated and because it is so thin it heats to the point of oxidation. Other tube structures nearby would be heated, of course, but that falls off as r^2 and because they have greater mass and the heating is both rapid and of short duration, they are not heated to the point of vaporization.
Yes, of course.
My question is with a known good tube how much reduction is seen in terms of ions (manifested as grid current) after re-flashing. It would seem that the gasses liberated through normal operation could be consumed, restoring the tube to a somewhat newish state. But that depends, of course, upon how much getter material remains active and able to bind with gasses.
The gettering (the "splash") is never not active unless it has fully reacted. Heating the ring will do nothing. Even if there is non splashed gettering leftover hanging on the getter ring, it will also be good active material even though it isn't on the glass. The splashing is just to create a larger surface area for the splash to have exposure to. The splash is active for the life of the tube unless consumed by reaction.
This is not talking about the ring, but about the getter itself which is deposited on the inside of the tube.
All tubes have gas. Every Single One. The issue is, of course, how much.
The metals used in tube construction all have absorbed or adsorbed gasses to some extent, and while a great deal of work was done in prepping the metals to reduce this, it could never be eliminated. The mica also contains oxygen. Heat, and time, cause these gasses to emerge. That, in turn, causes grid contamination and ions.
These liberated gasses will generally not react with the getter, except on the surface which has been passivated with an oxide (or other contaminant) layer after initial activiation, and then either ceases to react, or only slowly reacts, after the initial reaction has occurred.
I've read reports that induction heating, like this, can reactivate the getter and improve or rejuvenate a tube which is not suffering from a leak, but rather has absorbed gasses emerge and also has had the alkali oxides in the cathode decompose from heat and liberate oxygen.
I just don't know how true those reports are because they are somewhat lacking in statistical data.
The getter as applied during manufacturing is not active, and must be heated to activate it. Otherwise it would just be sputtered on or painted on and there's be no need to use a ring from which to vaporize it.
Yes, surface area is key. But the surface tends to be passivated, at least according to the reports I've read and the data we have on grid ions, which generally are gasses. So we have some proof that if the getter continued to react these ions would not be present.
Again, the ions are present only in very small quantities, but they do have an effect on the grid.
The gettering is heated for its application so that it can splash like hot solder. It will react hot or cold. There is no activation, only the splashing. The material is protected before and during assembly and starts doing its job as soon as it's splashed.
Last edited:
now, seems we are entering the academic discussions.
in my country, tube pumped down with induction heating to the internal structure and "over voltage cathode baking", after the pressure low enough, the tube sealed off. then again fire the getter in short period to get higher vacuum.
during the getter flashing, different types of gas molecule stick with the vaporized Barium and emended into the Barium mirror. after the getter flashing, Barium mirror almost only reacted to the oxygen.
when the tube in first use, the pressure first increase (mainly H2 H2O CO2 CO N2) in first tens hours, some times the pressure increased 1 decade or more, then decrease and almost stopped at certain (low) level ( mainly H2 Ar), the final pressure for a operating tube some time could lower than the pressure just after the getter flashing (for healthy and good quality tube only, most tube not).
after long period of storage the some of the tubes has its pressure increased due to slow leakage. if you can find the transparent spot on the mirror, the vacuum is for sure "bad", sometime 3 decade or more higher than after getter flash.
so the answer for you question,
healthy tube under normal using --> do not re-flash getter.
tube after long time storage --> if the mirror is perfect --> do not re-flash getter.
tube after long time storage --> if the mirror is not perfect --> re-flash getter
tube have blue light in the space between cathode and plate (please look though the plate window) --> re-flash getter
tube have blue light on the glass surface --> do not re-flash getter.
my gu-29 glow like mercury rectifier --> re-flash getter before you use it.
the major gas source (after 100h + use) is the mica and glass. and the gas release from them increased sharply when the temperature close to the decomposition temperature or soft temperature. many transmitting tube do not use mica at all and has hard glass even quartz envelope. oxygen from the oxide coated cathode is absorbed by Barium mirror.
I mean change is weak, no more than 1%.
induction heating work much better for Magnetic metal. the getter holder is made from coated steel in most case.
it's vacuum, anything has vapor pressure right ? higher temperature higher vapor pressure, and higher "transportation".
again, if the mirror is perfect, do not re-flash getter. do it only with glow tubes and tubes with bad mirror.
flash getter cannot renew the cathode emission ability. but it can prevent / reduce further damage.
This is quite interesting to me. Over a year ago I built a simple Royer type power oscillator to play around with wireless power and see if it would really be any use. In this form it won't but that's another subject.
I did do some induction heating experiments with it that were very fun and successful. I heated the getter in a 6AL5 for a laugh and it worked, as in it glowed. I didn't do any measurements, probably no point on a diode valve.
Anyway, I have a some 13E1. One of them is extremely crusty, cathode hotspot and vast grid current is what I scribbled on it when I tested it.
Now, hmmmmm. I'm thinking that if any valve could be resurrected and real measurements made it would be this guy. 35mA/V and 90W PDA. I don't think it is that gassy as there is no glow between electrodes just huge grid current and the cathode hot spot at 400V anode in triode at almost any bias point.
My wife is away tomorrow night so I think I will lash up the test rig again and test it. Then I will re-flash the getters that are about the size of a penny and test it again for grid current etc.
I do think that at least on larger valves re flashing the getter may work as I have seen a few people on YouTube re-use getters from larger valves such as PL519 in their own builds of valves. Pretty impressive glass work.
I also think in this case with this 13E1 that it won't fix it but I do believe I will be able to measure a difference. Be it better or worse.
I shall post some pictures and results this time tomorrow night.
Cheers
Matt.
I did do some induction heating experiments with it that were very fun and successful. I heated the getter in a 6AL5 for a laugh and it worked, as in it glowed. I didn't do any measurements, probably no point on a diode valve.
Anyway, I have a some 13E1. One of them is extremely crusty, cathode hotspot and vast grid current is what I scribbled on it when I tested it.
Now, hmmmmm. I'm thinking that if any valve could be resurrected and real measurements made it would be this guy. 35mA/V and 90W PDA. I don't think it is that gassy as there is no glow between electrodes just huge grid current and the cathode hot spot at 400V anode in triode at almost any bias point.
My wife is away tomorrow night so I think I will lash up the test rig again and test it. Then I will re-flash the getters that are about the size of a penny and test it again for grid current etc.
I do think that at least on larger valves re flashing the getter may work as I have seen a few people on YouTube re-use getters from larger valves such as PL519 in their own builds of valves. Pretty impressive glass work.
I also think in this case with this 13E1 that it won't fix it but I do believe I will be able to measure a difference. Be it better or worse.
I shall post some pictures and results this time tomorrow night.
Cheers
Matt.
I think the term "Vapor Deposition" is a better description of the getter activation action.
Sputtering is a term reserved for electrostatic deposition of material, as oposed to vapor deposition.
I believe "Splashed" implies a liquid volume thrown against a surface.
In this case the material is vaporized by induction heating, and deposited on the inside surface of the glass envelope as the vapor condenses in a partial pressure (no such thing as a complete vacuum?).
Certainly it remains active once deposited, but to what degree a surface layer could interfere with further operation is open to discussion since totally white deposits are seen in tubes which are defective indicating they remain so.
I have (or had) a 01A tube which was gassy. If I can find it, I will build an induction heater and try to see if I can recover the tube.
Sputtering is a term reserved for electrostatic deposition of material, as oposed to vapor deposition.
I believe "Splashed" implies a liquid volume thrown against a surface.
In this case the material is vaporized by induction heating, and deposited on the inside surface of the glass envelope as the vapor condenses in a partial pressure (no such thing as a complete vacuum?).
Certainly it remains active once deposited, but to what degree a surface layer could interfere with further operation is open to discussion since totally white deposits are seen in tubes which are defective indicating they remain so.
I have (or had) a 01A tube which was gassy. If I can find it, I will build an induction heater and try to see if I can recover the tube.
That is not correct.
Consider the chemistry.
To begin, the purpose of the getter is to chemically bind the oxygen or oxygen-containing molecules like carbon monoxide, and any other gasses like hydrogen or nitrogen. All gas should be bound. (The getter, however, cannot bind to krypton or argon because noble gasses have full orbitals and don't readily combine with other atoms.)
Any of the alkali metals could be used for this purpose, but barium is strongly preferred because it is inexpensive, already used in the tube, and won't poison the cathode. (The cathode is a typically either barium or a mix of barium and strontium so any barium that plates onto the cathode during getter flashing not only does no harm, but it might actually improve the cathode coating. This is not the case, however, for other tube structures.) Elemental barium is not stable in oxygen or water vapor (or nitrogen, carbon dioxide, etc.), of course, so the getter must either be stable in air (alloys) or be maintained in a vacuum or neutral atmosphere, until the tube was evacuated.
For the flashed getters (as opposed to getters deposited onto electrodes which are heated when the tube is in use), the use of an alloy (like barium plus aluminum) means a barium getter cannot be easily used unless the barium is activated after assembly. That's part of why the tube is flashed. The other reason is to evaporate the getter material to provide a thin layer with a very large surface area. No practical and cost-effective mechanism really exists to build a tube getter structure that functions as well as a thin layer of getter deposited on the inside of the envelope, which is why this technique is still used. Vaporization is easy and inexpensive and workable. The alternative would be some sort of thin shield around the tube's internal structures, which would cause all sorts of electrical and thermal issues. (Yes, there are patents for this. No, the technique generally is not used.)
Flashing further ensures the barium is nice and hot so it will quickly and readily bind with the oxygen as well as any carbon monoxide left from the heating of lubricants or oils on any metal parts. It also binds to hydrogen, nitrogen, and other gasses, but not argon or krypton.
Getters are typically made of a mixture of barium and aluminum (over iron), sometimes with magnesium added. Exotic getters, like those with tantalum, zirconium, etc. were more expensive and thus less common in a commodity item. The aluminum or magnesium protect the barium while being stable in air. But when hot these metals readily oxidize, consuming residual gas in the tube.
Slower vaporization of the getter causes the initial barium to oxidize, so the subsequent layers are deposited in a better vacuum. These layers don't oxidize, so they are shiny.
Rapid vaporization cause the vapor to be darker because the layers are a mix of metallic and oxidized barium. The darker version is a better getter.
So the getter doesn't splash, it completely vaporizes and deposits on the interior of the glass in a very thin film with very large area. Yes, it really gets this hot. If it splashed it would produce droplets and a highly variable layer, never an even film. Those droplets could also short out the tube if they cooled into balls and fell into the tube's internal structures. (The rest of the tube is cold as the induction heater does not evenly heat.) A vapor clearly lacks this droplet issue.
Ok. So that should explain how the getter works.
The surface of the getter should be porous and amorphous to readily allow gas to diffuse into it. But since a properly working tube may have gas (ions) and yet still have functioning getter material that could be oxidized, my question is what happens when the getter material is heated. This could, theoretically, drive off gas and make the problem worse, but at hte same time the temperature is low enough to promote oxidation but prevent breaking of the oxide bonds.
Last edited: