eHam

eHam Forums => Amplifiers => Topic started by: W1QJ on January 03, 2011, 09:28:21 AM



Title: How to getter a 3-500 tube
Post by: W1QJ on January 03, 2011, 09:28:21 AM
I am wondering if anyone that has encountered gassy 3-500z tubes has successfully managed to getter the tube and dissipate the gas and restore the tube to good use.  I have looked on the web for information on a tried and true method of gettering the gas in a tube but unable to find any information on going about doing it.  Of course there is plenty of information on "gassy tubes", but I am looking at possibly building a jig to getter the gas in these tubes I run across or even to supply the service to someone who may have a problem tube.  I think I have all the available parts to make such a jig, but i would need to know the particulars and the procedure to follow to successfully accomplish the task.  Any help out there?  Lou


Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 03, 2011, 10:11:18 AM
Lots of info on qrz.com. Check out http://forums.qrz.com/showpost.php?p=1991563&postcount=9 and the link http://www.w8ji.com/vacuum_tubes_and_vaccum_tube_failures.htm.

"Under some conditions a glass tube can be restored to operation by running low anode voltages and positive bias on the grid. This will sometimes allow full operating anode temperatures to be reached, and the tube can be "cooked" for several hours. I've had about a 50% success rate restoring old 3-500Z's that have sat for years without use. Even though they initially arced severely at full voltage, by cooking them at low voltage and positive grid bias to show anode color vacuum was restored."


Title: RE: How to getter a 3-500 tube
Post by: W1QJ on January 03, 2011, 11:16:52 AM
Thanks Mike, I have read that on the W8JI site.  I am just wondering if someone has more specifics.  Like what would be an appropriate "lower anode voltage"?  2kv?  How much bias at the specified "low anode voltage"?  Cook the tube  how long?  I am looking to make up a getter jig if you will.  I suppose I would need a glitch resistor and perhaps a fuse.  Tom claims he had 50% success, that's not too bad. 


Title: RE: How to getter a 3-500 tube
Post by: W8JX on January 03, 2011, 11:41:20 AM
Basically you need to get tube hot. You do not want to baby 3-500's as they need to glow dull to medium red when running. The Getter function of tube works best when tube is hot. BTW, this is not a problem at all for this tube and is "normal" operation for its graphic anode design. The coats on it need to get real hot to work as a getter.


Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 03, 2011, 12:40:56 PM
... I am just wondering if someone has more specifics.  Like what would be an appropriate "lower anode voltage"?...

The first link to qrz.com explains that.


Title: RE: How to getter a 3-500 tube
Post by: KH6AQ on January 03, 2011, 12:54:39 PM
3-500Z gettering parameters:

Plate at 500 volts
Grid at +60 volts
Plate current will be about 700 mA
Grid current will be about 200 mA

How long to getter the tube? If you have a hipotter you might hipot the tube Plate-to-grid periodically and getter it until it holds off 10 kV. It will not harm the tube to getter it for days.


Title: RE: How to getter a 3-500 tube
Post by: W8JX on January 03, 2011, 01:46:50 PM
I think what is being misconstrued here is that there is really 2 getter functions of a tube. There is the one primarily used when tube is first built and the secondary in the tube in question is were anode/plate is usually coated with a zirconium alloy of sorts that oxidies when it is heated up and in order to oxidizes it needs air to do it. (it gives it a grayish color) If a 3-500z is run properly and not babied it will heat anode enough that it will keep tube gas free barring a major leak. As long as tubes are not arcing over there is no need to try to do a primary or OEM get to tube as was done when it was built. Some get nervous when they see plates glow on a tube (and they should if it is a sweep tube or a 811) but it is quite normal for a 3-500z to glow dull to medium red and promotes a long life.   


Title: RE: How to getter a 3-500 tube
Post by: W1QJ on January 03, 2011, 03:43:53 PM
The info provided by WX7G is what I was looking for.  The plate voltage at 500 volts isn't enough to cause the gas to arc.  So I assume the tube plate will get real hot, getter the gas and then when put back into service it should not arc.  Is this a fair assumption?  Thanks for the information Dave, that was what I wanted to know.  Lou


Title: RE: How to getter a 3-500 tube
Post by: KH6AQ on January 03, 2011, 04:47:50 PM
Yes adjust the grid drive and/or plate voltage so the plate glows a dull red. It's not critical. A fan blowing some air on the tube is probably a good idea. I would hesitate to check an old tube in some amps that can be damaged by a tube arc.

This datasheet will show you what the tube needs: http://www.umich.edu/~umarc/station/docs/3-500z.pdf


Title: RE: How to getter a 3-500 tube
Post by: KM3F on January 03, 2011, 11:28:18 PM
As additional info, you need to know that gas levels in a tube can get so high that the gas ionizes and begins to conduct like a neon and will arc with very high voltages applied.
When this happens the number 1 grid has no control of electron flow so a run-away condition devlopes in a hurry. (arc)
Tubes that gas up to this extent usually cannot be recovered.
Also you may see some tubes that glow with a vivid blue on the inter glass surface.
This usually is not gas related but some contamination that glows when electrons strike the outer walls.  The tube may work fine with no ill effects otherwise..


Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 04, 2011, 02:21:57 AM
... it is quite normal for a 3-500z to glow dull to medium red and promotes a long life.

Correct. I would add, for additional emphasis, that it is not only normal, but ESSENTIAL. :-)

I was surprised that just 500 volts would work, since W8JI mentioned 1 to 1.5 kV in the link I provided above.

Something in the back of my mind tells me that after that filament is lit, you would want to get the plate good and red just as fast as possible. It seems to me that with air in the tube, that filament is living on borrowed time until the gettering is nearly complete.


Title: RE: How to getter a 3-500 tube
Post by: W1QJ on January 04, 2011, 05:39:29 AM
I have decided to build up a "getter jig", I will be able to adjust the HV from 0 to 1.5KV and I will have an adjustable dc supply for bias up to about 100vdc and of course filament voltage.  Someone mentioned a tube having too much gas and can not be restored.  Case in point, while checking the 3-500's I have, I very carefully installed them in an SB-220 amp with a variac on the line voltage.  I installed a known good tube in one hole and an unknown tube in the other.  Put the meter in the grid position without the HV hooked up.  Turn the amp switch on and brought the line voltage up slowly.  Without too much line voltage I noticed the grid meter going negative.  I knew there was problems right away!  BTW this is how most grid meters in SB-220 blow out!!  Although looking at the grid meter as I did indicated a bad tube right away, the small amount of current needed to make me know there was problems would not harm the meter.  Assuming you bought a used SB-220, installed the tubes, had the meter in grid position, turned it on with a filament to grid short or very gassy tube , POP, there goes your meter! (most likely).  I recommend turning the amp on in the HV position, if the tube is bad, you will know it!! And you will save your meter.  This tube turned out to have a short and I was able to read it with an ohm meter.  I hooked it up to a car battery and I saw a flash of light inside and then the short was cleared.  But when full HV was applied it appeared OK, no blue color.  I thought the tube was saved.  Then once the antenna relay was engaged and drive was applied, there was small arcs of blue gas.  SO this tube has some gas in it.  Interestingly several other tubes had gas at different levels.  Another tube had so much gas that about 800vdc ignited the gas and blew the grid grounding resistor I use like a fuse instead of the choke.  That tube most have major gas!!  I doubt that one can be rehabilitated?  Anyway, I recommend to anyone who has an amp to get a variac.  It is your best friend!  Not so sure when I will find time to build the jig, but it is something I want to do.  Tube arcs ruin lots of things in amps, meters, zener bias diodes, and grid shunt resistors.  With grids lifted above ground on some amps, grid caps and grid chokes burn up too.  Firing up an amp on a variac can save the "big bang" which does lots of damage.  Also, the variac will tell you if you have a shorted diode or cap in the PS right away.  The HV will not come up quick and there will be some humming of the variac pretty quickly.  You can almost diagnose any problem in the power supply and bad tubes within seconds of seeing what happens as the HV begins to come up on the variac, sparing the big bangs.  Since the variac is only for testing, a 5A variac for such use is OK and they can be gotten very cheap.  $20.00 at a hamfest is common for a 5a  variac.  Happy New Year


Title: RE: How to getter a 3-500 tube
Post by: W8JI on January 04, 2011, 05:40:49 AM
You are looking for a cookie cutter answer when there is none. There is no set voltage, except the voltage has to be lower than the ionization voltage of the gas. This typically is a few thousand volts, anode to grid, in a recoverable tube.

While low voltages can be used, low voltage requires more grid current (perhaps a few hundred mA) and a higher current anode supply. There is nothing wrong with that if you have a 40-80 volt 1/2 amp supply for the grid and a 600-800 mA 400-600 volt HV supply handy.

I use higher voltage because things are easier for me to manage. The AL80B has a tap on the transformer secondary that allows reduced voltage, and it is easier for me to come up with a variable supply that can be run up to 30 volts at 100 mA than it is something higher.

Plus the tube absolutely MUST have good forced air, so a higher speed fan in an AL80B chassis makes a nice fixture. Remember the tube filament is producing 75 watts of heat along with the 300 watts or so in the anode, and unlike amateur service this is continuous.

Asking for a set of operating conditions is counterproductive, because the best operating conditions would depend on what you use for a fixture. The only important thing is tube ratings are not exceeded, and the supplies you have will handle the load. As I said, and it is worth repeating, if a tube has a breakdown below a few thousand volts it almost certainly will not be recoverable.

A good 3-500Z, at minimum, should hold off three to four times the expected dc plate voltage. This is because in normal properly loaded operation peak anode voltage would be about twice the dc anode voltage. If the amp is mistuned or an excessive drive peak comes along, it is easy to go well beyond that 2 times value.

73 Tom


Title: RE: How to getter a 3-500 tube
Post by: W8JI on January 04, 2011, 05:52:37 AM
I have decided to build up a "getter jig", I will be able to adjust the HV from 0 to 1.5KV and I will have an adjustable dc supply for bias up to about 100vdc and of course filament voltage.  Someone mentioned a tube having too much gas and can not be restored.  Case in point, while checking the 3-500's I have, I very carefully installed them in an SB-220 amp with a variac on the line voltage.  I installed a known good tube in one hole and an unknown tube in the other.  Put the meter in the grid position without the HV hooked up.  Turn the amp switch on and brought the line voltage up slowly.  Without too much line voltage I noticed the grid meter going negative.  I knew there was problems right away! 

If you see stable negative grid current right away on a slow ramp up, you have a grid-to-cathode short and the tube is junk.

A gassy tube that is recoverable will have an uncontrolled sudden avalanche of current that occurs at some voltage generally up above a few kilovolts.

To protect the meters, a single diode is needed to clamp the negative power supply rail to the chassis.

A filament to grid short will not hurt the meters in an unmodified SB220. About all a filament to grid short can do is make the tube draw current when the amp relay is not closed. If the relay is closed some voltage less than 2.5 volts is applied to the path from grid to chassis, up through the grid meter, through the zener, and back to the filament CT. The very low voltage (always less than 2.5 volts) and the total resistance of that path (plus threshold of the zener) is virtually always enough to limit current to safe values.

Meter destruction comes from an anode to grid arc. This is generally either the anode falling over into the grid from bad welds, or gas in the tube going plasma.

73 Tom




Title: RE: How to getter a 3-500 tube
Post by: W1QJ on January 04, 2011, 06:15:59 AM
OK, if I install a tube that shows negative grid current with no HV on the plate, the tube has a short and is junk, Right?  Throw the tube away so you don't come back 5 years later and sell it or install it in an amp. Right?
 If a tube does not show negative grid current with HV at any level up to full operating HV, AND shows no blue gas color at HV idle, then you begin to drive the tube with RF and it arcs blue, What is the fate of this tube?  Can this one go in the getter jig?

I am trying to determine the parameters of a "junk" tube and a tube that might be able to be saved. 

Question:  I would think that a tube that shows negative grid current with no HV with just a small amount of filament voltage (bringing up with a variac) would have a pretty hefty short that should be able to be measured on an ohmeter.  One tube I did measure a short indeed, the other I did not.  What is conducting the current in that tube between filament and grid that is not detectable?


Title: RE: How to getter a 3-500 tube
Post by: W8JI on January 04, 2011, 07:37:49 AM
OK, if I install a tube that shows negative grid current with no HV on the plate, the tube has a short and is junk, Right?  Throw the tube away so you don't come back 5 years later and sell it or install it in an amp. Right?

Right. If it shows current with little or no HV in a good amplifier the tube is junk.

Quote
If a tube does not show negative grid current with HV at any level up to full operating HV, AND shows no blue gas color at HV idle, then you begin to drive the tube with RF and it arcs blue, What is the fate of this tube?  Can this one go in the getter jig?


I've sucessfully recovered tubes that arc or flash over at 2000 volts, but the odds of them staying good or even recovering are a lot less than tubes that only arc or flash at 4000 volts or more. I never ran statistics because I've only looked at 20-30 tubes and the sample is too small, but I can't recall ever being sucessful with one that failed below a few thousand volts.
Remember it is the PEAK voltage that is the problem, not the dc voltage. When RF is applied peak voltage can be much higher than dc. Look at the voltages here:

http://www.w8ji.com/demonstation.htm

Quote
I am trying to determine the parameters of a "junk" tube and a tube that might be able to be saved. 


I've had tubes go right back to being bad, even those that initially high potted to 4kV or more, but I have a lot less luck with tubes below 2-3 kV.


Quote
Question:  I would think that a tube that shows negative grid current with no HV with just a small amount of filament voltage (bringing up with a variac) would have a pretty hefty short that should be able to be measured on an ohmeter.  One tube I did measure a short indeed, the other I did not.  What is conducting the current in that tube between filament and grid that is not detectable?

Heat moves the filament slightly. I test for grid-cathode with reverse bias current on a hot tube. My bias supply is 1000 volts maximum and current limited to 20 mA. An ohmmeter not only has too low of a voltage, it also can't safely be used with full filament voltage for some meters. This is because there might be half of filament voltage put across the meter.

I also bang on the tube while watching for faults.

The test fixture I built for 8877's back when Eimac was having problems with H-K shorts ran the filament voltage up and down every few minutes, and had a fairly high voltage negative bias on the grids through a relay. The relay would latch if a tube ever conducted. We ran tubes for at 24 hours that way to cull out the bad ones. The tubes required a fan even with just filament power. It tested several tubes at once. :-)


Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 04, 2011, 07:51:44 AM
... I am trying to determine the parameters of a "junk" tube and a tube that might be able to be saved.  ...

From reading this post and dozens of others like it, I'm starting to come to the conclusion that the parameters of a junk tube is, if it has "3-500Z" printed on the glass. Geez-o-petes. What did Eimac do wrong? All glass power amp tubes aren't that bad, are they?


Title: RE: How to getter a 3-500 tube
Post by: AA4HA on January 04, 2011, 08:17:27 AM
From reading this post and dozens of others like it, I'm starting to come to the conclusion that the parameters of a junk tube is, if it has "3-500Z" printed on the glass. Geez-o-petes. What did Eimac do wrong? All glass power amp tubes aren't that bad, are they?

I think you jest, the 3-500's are just a very old design that has been out of manufacturing for quite a few years. The folks who are reviving 3-500's are trying to extend out the life of amplifiers that are based around that particular tube. As quantities of recoverable tubes continue to decline then the measures that folks will go through to revive the near dead will get more extreme. Eventually there will be one ham who owns the last 3-500's who relearns the art of glassblowing so they can rebuild tubes from the inside. (or really tiny people in space suits who can replace a burnt out filament).

Since it is glass there is always a minute amount of gas that will leak in through the glass itself or around tube seals. (glass really is a super high viscosity liquid, not a solid).

I would go so far to say that the two markets that have continued to keep ceramic tubes alive are in industrial induction heating applications and MRI machines. If Eimac was just making tubes for hams they could have five employees working out of a barn. Industrial/Medical customers are their bread and butter.

Tisha Hayes


Title: RE: How to getter a 3-500 tube
Post by: WB2WIK on January 04, 2011, 08:58:12 AM
... I am trying to determine the parameters of a "junk" tube and a tube that might be able to be saved.  ...

From reading this post and dozens of others like it, I'm starting to come to the conclusion that the parameters of a junk tube is, if it has "3-500Z" printed on the glass. Geez-o-petes. What did Eimac do wrong? All glass power amp tubes aren't that bad, are they?

Eimac hasn't made this tube in a very long time.  Others still do.

I think it's a great tube, if you buy a new one and use it a lot.  They can last 10+ years in daily service, yielding an annual cost of $17.50 for a tube that can produce a kilowatt output on HF.  Not a bad deal at all.

The "problems" are mostly with storage life.


Title: RE: How to getter a 3-500 tube
Post by: W1QJ on January 04, 2011, 09:42:32 AM
You Know what guys?  I think you talked me out of the getter jig.  It seems like the success rate is low and the problem most likely will return again in time.  Problem is I have bought several hundred SB-220 amps over the past years and unfortunately the sellers have not been very honest.  You pay a price that should reflect a working set of tubes and you get one or both that are bad.  I could suck this up if I was buying one amp and needed to buy a set of tubes that would last me 10 years, but with what I do, if I get a bad tube or tubes, I am out.  I can't tell you how many of the amps I bought that were "AOK" had probelms!  More than I care to imagine.  Thankfully I can fix any problem that comes down the pike except for the bad tubes.  Not to mention how many of these amps had bad band switches.  Thank God that does not affect me, but I fear for the poor guy who is looking for a good working HF amp and buys it.  After several hundred SB-220's in out of here I estimate 7 out of 10 amps have a burned band switch contact.  usually 15 or 10 or both, but I have seen eveyone along the way.  So, got a bad SB-220 you want to unload, let me know I'll take it off your hands.  Thanks Tom and the rest for the good info on this topic.  Lou


Title: RE: How to getter a 3-500 tube
Post by: KH6AQ on January 04, 2011, 09:48:45 AM
Transistors never need to be gettered.


Title: RE: How to getter a 3-500 tube
Post by: W1QJ on January 04, 2011, 11:36:37 AM
I guess some day I'll just have to give in to solid state amps like I did with transceivers.  It may be sooner than I think. ;D


Title: RE: How to getter a 3-500 tube
Post by: G3RZP on January 04, 2011, 02:23:19 PM
>Transistors never need to be gettered.<

Doesn't mean that they are good, or necessarily, that they will stay good.

30 + years in the semiconductor industry doesn't convince me....even having done space grade stuff.


Title: RE: How to getter a 3-500 tube
Post by: W8JX on January 04, 2011, 03:12:10 PM
Transistors never need to be gettered.

Tubes generally never need to be either if they are run properly.


Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 04, 2011, 05:27:06 PM
I think it's a great tube, if you buy a new one and use it a lot.  They can last 10+ years in daily service, yielding an annual cost of $17.50 for a tube that can produce a kilowatt output on HF.  Not a bad deal at all.

The "problems" are mostly with storage life.

I agree 100%. I wasn't very clear when I said what I did. However, not everyone can avoid storing them, and that's the big "if" here.

My question is whether tubes like the 3-1000Z, 4-1000A, 4-400A, 833A, 833C, and other glass tubes have a storage problem like that -- a pin seal leakage problem -- whether made by Eimac or somebody else.


Title: RE: How to getter a 3-500 tube
Post by: K6AER on January 04, 2011, 05:43:39 PM
Three factories in China are making 3-500Z and ZG's. The need for these tubes are fed primary by the amateur radio market. Most manufactures of industrial equipment use ceramic tubes or have gone solid state.

MRI's stopped using tubes 6-7 years ago. The days of cheep pulls are long gone.

Aside from rotating glass tubes each year so they can be gettered, the proper use of the tubes has nothing to do with the glass seal leaking or not. That is a function of manufacturing luck and some skill. I have scene EIMAC's fail that were just manufactured and Chi Com tubes run forever.

The most important decision is to buy the tube from a supplier with a warranty.


Title: RE: How to getter a 3-500 tube
Post by: W8JX on January 04, 2011, 05:52:00 PM
the proper use of the tubes has nothing to do with the glass seal leaking or not.

While you cannot prevent leaks (and all tubes "leak" to some degree) you can mitigate the effect of it by making sure tubes get to proper operating temperatures the absorb what gasses may be present.


Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 04, 2011, 06:20:04 PM
Let me ask this question again, in another way.

MANY commercial transmitters in the USA use 4-400As. This is not an uncommon tube for amateurs to use. Do 4-400A's (or similar tubes in the same power range as the 3-500Z) have the same storage problem as the 3-500Z or not?




Title: RE: How to getter a 3-500 tube
Post by: KM3F on January 04, 2011, 09:49:54 PM
Doing this is a good learning tool and experience but to expect to recover any tubes with any degree of certainty for others is a bit of over expectation and risky for friendship and reputation.
You must be awhere of max grid 1 dissipation or you will melt or distort the grid structure.
The plate anode needs at least 400 watts of dissipation to even begin to show color for gettering action on the 3-400/500 tubes.
You don't have to draw grid current unless there is some special reason or you want to see if grid control is proper.
About the only special time you may need to attempt to draw grid current is to burn off any contamination that causes the grid to heat and begin to lose negative bias control over plate current and causing a run-away condition.
At zero bias or grounded grid, any combination of voltage and current that will cause the anode to dissipate 400 watts or more is enough.
Then the time and activation cycling has to be played with for each tube in order to see if a recovery is becoming worthwhile.
Good luck.


Title: RE: How to getter a 3-500 tube
Post by: W8JI on January 05, 2011, 07:26:56 AM
Let me ask this question again, in another way.

MANY commercial transmitters in the USA use 4-400As. This is not an uncommon tube for amateurs to use. Do 4-400A's (or similar tubes in the same power range as the 3-500Z) have the same storage problem as the 3-500Z or not?




Yes, they all do. Some problems come from the kovar used to seal the pin to the glass. It is subject to deterioration from moisture.

I remember back in the prime of Eimac. Their engineers actually visited larger OEM's. Conversations during visits were worth their weight in gold. Jim Aurand and the senior engineer from Eimac Salt Lake, where the glass tubes and 8877 were made, told me how RCA overkilled the dissipation of a driver in a BC transmitter and was having life problems with tubes. They said they dropped back from a 4-400 to something smaller (perhaps a 4-250, I can't recall) and the life problems went away. This is because the anode started to show slight color.

We also had deep conversations about the grids, filaments, and other things. This was because I wanted to know exactly how hard we could run the 3-500Z in the SB1000/AL80 amp, and what life or damage we could expect.

While grid current in the gold plated grids of 3CX800 and 8877 tubes is critical, their senior engineer was very clear that grid current in the 3-500 was not a major concern except for proper tuning. He said the grid dissipation limit was set in tests by secondary emission, and not by heat failure or life.  He explained the test method.

Since this came from the actual tube design engineers, I go by what they told me. (I also saw they had hand-drawn cartoons about nichrome parasitic suppressors on the walls in engineering, so they paid attention to the outside world. )

73 Tom



Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 05, 2011, 10:00:45 AM
So that's it. Thanks, Tom! :-)

I thought 833A's and a few other glass tubes didn't have that problem. I stand corrected.

That's funny about Measures' suppressors. I'd have liked to have seen those cartoons.


Title: RE: How to getter a 3-500 tube
Post by: KH6AQ on January 05, 2011, 04:38:29 PM
At zero bias or grounded grid, any combination of voltage and current that will cause the anode to dissipate 400 watts or more is enough.

With zero volts on the grid a 3-500 needs 3 kV on the anode to dissipate 400 watts. This voltage might be too high for a gassy tube.


Title: RE: How to getter a 3-500 tube
Post by: KM3F on January 05, 2011, 05:24:18 PM
Here are more things to keep in mind.
If the grid has an issue, it likely is due to being over driven such that the grid structure has sagged out of place and ended up very close to the filament.
The effects of this condition can make the tube sensitive to vibration and temperature changes during operation.
Be awhere that switching a tube's filament on creates a physical shock to the filament that can make it actually move. This could result in a short with the grid that has been abused.
The basis for this movement is the filament resistance is very low when cold such that high inrush current occurs before the filament heats and it's resistance increases.
As well, a test of plate current (Ib) VS  negative Grid bias (Eg) may not track a known good tube because the grid or it's turns are out of place physically or even one or more grid turns had been burned out and lay at the bottom of the tube structure.
Leakage should be rare in this type structure due to how the elements are mounted to the stem unlike a small receiving tube that often suffers from leakage between the elements having a leakage path accross the mica structure that hold all the elements in there precision alignment.
All of these items are the basis of pass or rejection at the time of mfgering.
Good luck.


Title: RE: How to getter a 3-500 tube
Post by: W8JI on January 06, 2011, 05:22:03 AM
Here are more things to keep in mind.
If the grid has an issue, it likely is due to being over driven such that the grid structure has sagged out of place and ended up very close to the filament.
The effects of this condition can make the tube sensitive to vibration and temperature changes during operation.

Not likely in the 3-500Z because of grid materials and construction. Most grid-to-filament shorts are the result of manufacturing issues or physical shock. The grid is very rugged and pretty much immune to damage from exciter power. Not so with grids in oxide cathode tubes like the  8877 or 3CX800A7, but in thoriated tungsten tubes are different.

Quote
Be awhere that switching a tube's filament on creates a physical shock to the filament that can make it actually move. This could result in a short with the grid that has been abused.
The basis for this movement is the filament resistance is very low when cold such that high inrush current occurs before the filament heats and it's resistance increases.



This is largely another non-issue with the 3-500. We'd be hard pressed to find an amplifier that stresses the filament to the point of moving it. The real dangers seem to be repeated heating and cooling as tube dissipation moves from a cold tube to a few hundred watts and back again over and over and over, not the turn-on shock.

The only amp I ever saw that probably had turn-on shock failure was one that had the filaments turned off and on every single transmission. In an effort to extend tube life someone actually wired the filament to his TR relay system with a little time hang delay system. Series filaments can be bad news too.

I had long discussions with Eimac's senior engineer about filament inrush. He told me when they evaluated that in the lab, they strapped the filament pins of 3-500Z's directly to 200 ampere transformers and after hundreds of hard cycles could not crate a problem. He pretty much dismissed the idea that inrush was a concern.

More than that, if we measure the inrush of amplifiers, we are very hard pressed to find any that go much beyond a 2:1 current ratio on start up. Heath took a hard hit back in the 70's for inrush in amps. This was part of hysteria started by people who just couldn't take the fact that tubes go bad, and were assigning blame to inrush. Articles were written and published about correcting inrush without actually measuring inrush. The SB220 barely makes 2x running current on the first few cycles of start, and then current rolls off rapidly.

The worse tubes and systems I see for inrush are the slow heating oxide cathode tubes, and even they generally are not too bad unless the filaments are in series or the filament transformer is far too large in current capacity and the filament wiring is heavy.

One person attributes filament movement to magnetic fields, but if forces from magnetic flux are calculated they are on the order of only a few grams in smaller tubes like the 3-500 or 3-1000!!!

Quote
As well, a test of plate current (Ib) VS  negative Grid bias (Eg) may not track a known good tube because the grid or it's turns are out of place physically or even one or more grid turns had been burned out and lay at the bottom of the tube structure.

True.

Quote
Leakage should be rare in this type structure due to how the elements are mounted to the stem unlike a small receiving tube that often suffers from leakage between the elements having a leakage path across the mica structure that hold all the elements in there precision alignment.

Ceramic also, not just mica. Chinese tubes light up the ceramic from impurities:

http://www.w8ji.com/572b_problems.htm

73 Tom


Title: RE: How to getter a 3-500 tube
Post by: W5WSS on January 07, 2011, 07:51:06 AM
My old amplifier uses two Eimac 4-400a. Due to circumstances beyond my control, it was shut down and stored. The Johnson Invader 2000 revamped into an amplifier was shut off for 12 years. The separate power supply is in original condition.uses mercury vapor rectifiers! I admit I expected immediate problems but was amazed after just "turning it on" everything started up. Amplified properly.


Title: RE: How to getter a 3-500 tube
Post by: K9PU on January 07, 2011, 02:20:11 PM
Consider placing the tube in a hard vacuum for an extended period of time (likely months.)  This, in theory, should allow gas to exit the tube.  High temperature inside the tube should allow for a faster exit of gas within the tube.  A hard vacuum is hard to get though, more than just a vacuum pump. 

If you have a means of measuring the internal vacuum of the tube you can monitor its improvement.

If the tube has a leak this may be for naught.  Of course you could leave it in the vacuum.

Good luck with this.

Scott


Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 07, 2011, 11:25:08 PM

What about using externally applied RF induction heating to get the plate red hot? Practical or not, what are the chances?

I don't fully understand the physics of the gettering process. I don't have a problem with the concept that the oxygen in the tube is gettered; I think of this as an oxidation process. However, nitrogen is an inert gas. How can it combine with the gettering material? It must, though, because tubes that don't sit idle for many years don't have the problem.


Title: RE: How to getter a 3-500 tube
Post by: W8JI on January 08, 2011, 02:44:51 AM

What about using externally applied RF induction heating to get the plate red hot? Practical or not, what are the chances?

I don't fully understand the physics of the gettering process. I don't have a problem with the concept that the oxygen in the tube is gettered; I think of this as an oxidation process. However, nitrogen is an inert gas. How can it combine with the gettering material? It must, though, because tubes that don't sit idle for many years don't have the problem.


If the tube has so much air it can't even be lit up normally, it is almost certainly done for anyway. Apply a reasonable but low anode voltage, some positive grid voltage, have good cooling in place, and heat the tube.

Many tubes can be gettered by running the HV lower than full (like the SB220 on CW, or the AL80B on the lower voltage HV tap) and drive the amp with light drive jus enough to show color with the anode out of resonance. The only thing you need to watch is tube seal temperature, because the SB 220 only really can cool about 200-300 watts of dissipation per tube with the stock fan.

Anything more than that is not worth it. It's not a $3000 tube....yet.

73 Tom


Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 08, 2011, 11:39:50 PM
... the SB 220 only really can cool about 200-300 watts of dissipation per tube with the stock fan.

Oh really? Is that the reason for the lower anode voltage "CW" switch position, or is the transformer the limiting factor?


Title: RE: How to getter a 3-500 tube
Post by: W8JI on January 09, 2011, 12:24:44 AM
... the SB 220 only really can cool about 200-300 watts of dissipation per tube with the stock fan.

Oh really? Is that the reason for the lower anode voltage "CW" switch position, or is the transformer the limiting factor?

The SB220 was designed when the legal amateur power limit was 1000 watts DC input minus the driver stage power. That's about 600 watts output carrier.

As such, there was no need to cool the tubes or amp much beyond that. That's why people unsolder the pins in the tubes when they try to run a kilowatt output.

The tap on the transformer was to allow tuning and operating the amp at legal CW power with about 70% anode voltage, and then when flipping to SSB tuning remains the same because plate voltage and plate current come up at the same ratio. This approximately doubles input power on peaks while average power remains well below 1 kW. (It would be illegal to have the plate current meter ever swing past the amount that allows 1000 watts on the highest voice peak.)

It was a entirely different world before the 80's.




Title: RE: How to getter a 3-500 tube
Post by: W0BTU on January 09, 2011, 03:36:36 AM
...It was a entirely different world before the 80's.

Oh, yeah! Why didn't I remember that? :-)

Thanks, Tom.


Title: RE: How to getter a 3-500 tube
Post by: KM1H on January 14, 2011, 02:33:55 PM
I almost sprayed my afternoon coffee while reading Toms comment about Eimacs nichrome cartoons! When I called Eimac after that infamous QST article I was told that xxx was responsible for the destruction of more 8877's than all other causes combined. He is still in parasitic mode on forums.

As far as Eimac 3-500Z production problems Im aware of the late 80's seals problem where tubes went to air just sitting NIB for 2 years. I believe this was after production moved to SLC. Another SLC adventure was when they tried to save money and just spot welded the top support to the anode. These broke loose, singly or all, causing the anode to either lean over or actually contact the grid.  More rumors sold many more uneeded inrush limiters. In the SB-220 family they do protect the AC mains switch; turning on in the CW position does about the same.

Ive been using a much modified SB-220 chasssis rescued from a CBer to getter tantalum plate 3-400/500's, 4-250/400's, and PL-175's since the late 90's (as well as old school tubes such as 250TH, 304TL, etc on another jig). HV is around 900V and grid voltage is variable; Ive changed things around since the first version I described on the AMPS forum and elsewhere back then. Ive done a few hundred 3-500Z's since then as every amp I convert to 6M automatically gets the treatment even before I test them on the Hi-pot.  The success rate is higher than Tom as Im obviously doing it to many perfectly good tubes but its a habit I got into and it takes no real effort on my part and it gives me that warm fuzzy feeling ::) Some test beyond 12KV which makes for good old Class C DSB AM candidates.


Now this brings up another question: Can you degas a graphite plate by bringing it to a low color?  I understand that getting them into the orange range actually releases gas. The few used Chinese 3-500's Ive done with a before/after hi-pot have shown an improvement but I didnt go beyond a dull red for a couple of hours.

Carl
KM1H



Title: RE: How to getter a 3-500 tube
Post by: AA4HA on January 15, 2011, 09:11:23 AM
What about using externally applied RF induction heating to get the plate red hot? Practical or not, what are the chances?

I don't fully understand the physics of the gettering process. I don't have a problem with the concept that the oxygen in the tube is gettered; I think of this as an oxidation process. However, nitrogen is an inert gas. How can it combine with the gettering material? It must, though, because tubes that don't sit idle for many years don't have the problem.

Gettering is pretty simple when you understand the chemistry of the materials involved, the process of heating a getter for the first time and the conversion that occurs during the gettering process. You will also see the limitations of re-gettering a vacuum system.

Gettering materials are usually reactive, alkali metals in their pure form (no oxides present). The gettering component may be a coating on the inside of the tube, a little halo like metal structure or other odd looking components between the tube structure and the glass wall. Some tubes may make the anode out of a gettering material as well.

When a tube is manufactured they try to draw down the internal pressure as close as they can to being a hard vacuum. Sometimes a little bit of heat is added to encourage structures to outgas while the vacuum pump is still running. Once they get to a certain point the glass tube is heated, melted and twisted off. Some tubes have the characteristic glass dimple on the top of the glass envelope, others it may be in the base (like an ordinary light bulb).

Now you have a tube that has a pretty good vacuum but not quite perfect. For tubes with the little halo of metal (zirconium, barium, etc...) (pretty much any metal that has a name that ends in "ium") they may be put in an induction machine to heat that metallic component up to extraordinary temperatures. As the metal then cools, it sucks up oxygen molecules to form barium oxide, zirconium oxide, xxxium oxide...) For most tubes the gettering metal is a one time process at the point of manufacturing but some tubes can be re-gettered by heating them up again and letting the alkali earth metal reabsorb the trace gases in the tube.

You can see that there is a practical limit to this. The amount of alkali earth metal in the tube that can be converted into an oxide may be limited or sometimes by reheating a tube you can cause outgassing that ends up plating somewhere else. Also, gettering metal compounds may be mixed to draw down certain elements (oxygen, hydrogen, etc...) but may leave other gasses alone in the tube.

If you have a tube that is genuinely leaky and is losing vacuum you may get one or two shots of extending the life of a tube by re-gettering but evenutally the gettering will just give out completely as it is overwhelmed.

Some tubes may use a different process that is similar to gettering. This is similar to what is used in ultra-high vacuum systems and is called an "ion pump". (sputtering pump). It involves a cathode that is porous to gas and heat. I do not remember if the 3-500 is a conventioal alkali earth getter or if the anode may also be working as a sputtering pump when heated.

This is all +25 year old information from university physics classes and from doing some internship time at a few national laboratories.

Tisha Hayes


Title: RE: How to getter a 3-500 tube
Post by: W0BTU on August 12, 2011, 08:21:04 PM
Nice photo on page 152 in the September QST by W8JI, which powerfully illustrates why 3-500Zs need to be periodically gettered.


Title: RE: How to getter a 3-500 tube
Post by: AG6K on October 04, 2011, 03:34:18 PM
I think what is being misconstrued here is that there is really 2 getter functions of a tube. There is the one primarily used when tube is first built and the secondary in the tube in question is were anode/plate is usually coated with a zirconium alloy of sorts that oxidies when it is heated up and in order to oxidizes it needs air to do it. (it gives it a grayish color) If a 3-500z is run properly and not babied it will heat anode enough that it will keep tube gas free barring a major leak. As long as tubes are not arcing over there is no need to try to do a primary or OEM get to tube as was done when it was built. Some get nervous when they see plates glow on a tube (and they should if it is a sweep tube or a 811) but it is quite normal for a 3-500z to glow dull to medium red and promotes a long life.   

  good advice.  Congrats.


Title: RE: How to getter a 3-500 tube
Post by: N2EY on October 04, 2011, 05:25:00 PM
I do not remember if the 3-500 is a conventioal alkali earth getter or if the anode may also be working as a sputtering pump when heated.

IIRC:

3-500Z has a zirconium-coated anode which does the gettering.

833A has a zirconium-coated anode, plain old 833 has a tantalum anode.

The 4-xxxx series of glass tx tubes, from 4-65A to 4-1000A, have anode getters but I don't know if they did the same no A = tantalum, A = zirconium thing.

Such anode-getters must be hot to work; this is why it is a good idea to rotate which tubes are used if you have spares.

It is also why it can be a bad thing to run those particular tubes far below rating - say, 250 watts input to a tube capable of 1000+. While it might seem that doing so would give incredibly long life (as it can with receiving-type tubes), the actual result can be shortened life due to no gettering because the anode never gets hot enough.

---

There was an article in QST for October, 1936 describing a 1 kW input tx that used 8 Eimac 35T tubes plus an 80. 35T crystal oscillator, 35T buffer/multiplier, push-pull 35Ts in the final. 2500 volts 400 mA! (though they usually ran them at 1500 volts 250 mA - "only" 375 watts).

The other four 35Ts were used as grid-controlled rectifiers, which permitted keying the B+.

Wanna bet those 35Ts were well gettered? With the wide open breadboard construction, cooling and seeing your way around the shack wasn't a problem...

That the authors were W6UF and W6CHE may have influenced the tube choice.



73 de Jim, N2EY       


Title: RE: How to getter a 3-500 tube
Post by: AG6K on October 05, 2011, 05:23:10 AM
The info provided by WX7G is what I was looking for.  The plate voltage at 500 volts isn't enough to cause the gas to arc.  So I assume the tube plate will get real hot, getter the gas and then when put back into service it should not arc.  Is this a fair assumption?
 

  Since the leaky seal that allowed air to enter the envelope in the first place is still leaking, it is a fair assumption that air will continue to enter.  Rich, ag6k


Quote
Thanks for the information Dave, that was what I wanted to know.  Lou


Title: RE: How to getter a 3-500 tube
Post by: G3RZP on October 05, 2011, 05:51:07 AM
When a 35T dissipated 35 watts, you almost could see to read by the light! Same with some 24G. Great stuff, tantalum.


Title: RE: How to getter a 3-500 tube
Post by: AG6K on October 05, 2011, 09:42:45 AM
When a 35T dissipated 35 watts, you almost could see to read by the light! Same with some 24G. Great stuff, tantalum.

 Indeed.   Tantalum (Ta) has such a high melting point (2,996°C) that it can be used for filaments in light bulbs.   cheers,  Rich


Title: RE: How to getter a 3-500 tube
Post by: GM3SEK on November 07, 2011, 07:41:16 AM
I do not remember if the 3-500 is a convential alkali earth getter or if the anode may also be working as a sputtering pump when heated.

IIRC:

3-500Z has a zirconium-coated anode which does the gettering.


The 3-500Z actually has three different getters, operating at different temperatures in different parts of the structure.

The following procedure (copied from the AMPS archives) came directly from a former Eimac tube designer, who was with the company when they were still manufacturing glass tubes.


<quote>
*************************

If a tube was last used years ago, or was manufactured years ago, or is new, it's safest to assume it needs gettering. And the suggested routine is this:

Cooling is of course required throughout.

a.  Bring the filament up to rated voltage slowly.  For example, run the primary of the filament transformer through a light bulb, or similar to limit the shock of turning it on.

b.  Then after some time, bring the voltage to the specified value, and run it for about 100 hours.

c.  Then, apply the anode voltage through some current limiting means for a while, and, finally,

d.  Run the tube at a dissipation level near the specified limit for a few hours.

The reason for this is that the getter needs to experience a spectrum of temperatures, mopping up different species at different temperatures.  Furthermore, the lowly 3-500Z for example, has getters at the base of the grid and the base of the filament as well as sprayed on the anode.

While this conditioning schedule is seemingly over-complicated and ultra-conservative, it should result in no surprises when putting the tube into service.

Terms like "a few hours" are vague because different tubes require different treatment.  If we're "burning in" a monster tube that will cost us $50,000 to replace, we're going to be more conservative than with a $50 tube.

***************************
<end quote>


73 from Ian GM3SEK


Title: RE: How to getter a 3-500 tube
Post by: AG6K on November 07, 2011, 08:53:09 AM
I do not remember if the 3-500 is a convential alkali earth getter or if the anode may also be working as a sputtering pump when heated.

IIRC:

3-500Z has a zirconium-coated anode which does the gettering.


The 3-500Z actually

  a key-word, N.B..

Quote
has three different getters, operating at different temperatures in different parts of the structure.

The following procedure (copied from the AMPS archives) came directly from a former Eimac tube designer, who was with the company when they were still manufacturing glass tubes.

  Ian: what is the name of this tube designer?  tnx

Quote


<quote>
*************************

If a tube was last used years ago, or was manufactured years ago, or is new, it's safest to assume it needs gettering. And the suggested routine is this:

Cooling is of course required throughout.

a.  Bring the filament up to rated voltage slowly.  For example, run the primary of the filament transformer through a light bulb, or similar to limit the shock of turning it on.

b.  Then after some time, bring the voltage to the specified value, and run it for about 100 hours.

c.  Then, apply the anode voltage through some current limiting means for a while, and, finally,

d.  Run the tube at a dissipation level near the specified limit for a few hours.

  How was it  that this critical information got left off of not only Eimac's Tech Data sheets for the 3-500Z , but it was also left off of all other Eimac glass-envelope tube data sheets?

Quote
The reason for this is that the getter needs to experience a spectrum of temperatures, mopping up different species

  now there's a new one. 

Quote
at different temperatures.  Furthermore, the lowly 3-500Z for example, has getters at the base of the grid and the base of the filament as well as sprayed on the anode.

  I've not seen any other substance on the grid of a 3-500Z that I have autopsied. 

Quote
While this conditioning schedule is seemingly over-complicated and ultra-conservative,

  to me this important  revelation is a tad suspicious due to its tardiness.

Quote
it should result in no surprises when putting the tube into service.

Terms like "a few hours" are vague because different tubes require different treatment.  If we're "burning in" a monster tube that will cost us $50,000 to replace, we're going to be more conservative than with a $50 tube.

  "Monster" tubes have built-in ion pumps.  Rich, ag6k

Quote
***************************
<end quote>


73 from Ian GM3SEK