Rebuilt 15GP22's?
 

Picked this up from the A.R.F. Site.

Amongst other things they claim to be able to rebuild the 15GP22.

Anyone know about this place? I wouldn't mind having a 3KP4 rebuilt.

http://www.wmicronics.com/rebuilttubes.htm

Eric:

That can not be correct. Pete Deksnis has been involved in a 15GP22 rebuilding project for the last several years and they are just finding a way that might work. The rebuilding progress can be followed on his site. I am sure that if some company was already doing that he would have known. I can't wait until someone does find a way!

Steve

Ummm...why do you s'pose it would be listed along with the two early 21" round color tubes among "black & white picture tubes" they claim to rebuild? At first glance, less than credible to say the least.

No way, I'm afraid at this time. There have been many attempts to rebuild 15GP22s, and all have been unsuccessful. As was explained to me several years ago by Gino at MetroColor, the metal flange on the 15G will separate from the bell of the tube as soon as they try to heat it. They tried, with donor 15Gs from Ed Reitan, Gary Miller, and myself. And back in the late 80s, Gary Hough and I tried an experimental rebuild with Dunbar. We couldn't get the 15G gun, but we had two tubes that still had filament continuity re-evacuated. They lasted about a month. The best explanation of "why not?" came from the guys at MetroColor, who have been there since the 50s, and held government contracts and met all ISO guidelines. In other words, they know their tubes. To rebuild the tube, the oven will have to be modified to heat at a precise rate and at a precise temp. There are also many other problems.

Bottom line, be skeptical of anyone who claims they can do it. UNLESS it is one of the women who worked on the original assembly line for the 15GP22, LOL...

Charles

i call bs.
and the big problem with these tubes breaking in the oven is likely due to the complex mask parts and their different rated of expansion.
the mg seal is usually where they leak so no surprise they fail upon heating.

Here's a snippet from an article in the Cleveland Plain Dealer on the 50th anniversary of the CT-100. It quotes Steve McVoy discussing the flange/baking-temperature issue associated with the ongoing 15GP22 rebuild effort.

When the picture tubes were manufactured, air was removed at room temperature by a vacuum machine. The picture tube then was heated in a 700-degree oven designed to excite the remaining air molecules so that as many as possible could be pulled out with the vacuum machine.

With time, the adhesive fails and air leaks through the seam. Restorers are experimenting with a new adhesive that has potential to keep air out for a long time, but the adhesive degrades at 500 degrees. Putting it in a 700-degree oven would not work.

"Right now, we don't know why 700 degrees is the magic number," McVoy said. "Was that the perfect temperature to remove all the air, or will something less than 500 degrees be enough?

"If it works, then we could have lots of tubes," he said. "I have four tubes and six sets that could use them." In addition to the RCA CT-100 and Model 5 prototype, several other brands of TVs of that era used the 15-inch color picture tube.

And another from the CT-100 news & noticespage of my site.

09-05-2005...on the 15GP22 rebuild front, a new, custom-made cam that controls the slew rate of the baking temperature has been ordered. After a few more hurdles (probably most of them as yet unknown), the rebuild team will attempt an engineering evaluation run using a damaged 15GP22 before a bona fide attempt is made to bring a rebuildable tube back to life.

Simply put, a month is plenty of time to slop on a bead of sealant and dry it. If the tube survives baking for that long without the custom slue rate, old 15GP22's may soon be looking a lot brighter!

Wonder how they made th' damthings in the first place ? You mean back in prehistoric 19-fackin'-54 they did somethin' we can't do today, w/all our glorious tecknologically & such ?!? Of course, maybe they were a-feared of pissing off David Sarnoff...I hear he didn't exactly suffer fools gladly....-Sandy G.

If RCA originally designed these tubes, you would think there would be archives somewhere of the tube designs and how to manufacture them.

I would think it would be cheaper to manufacture brand new tubes than to break so many in existance trying to find a way to put a new phosphers and gun assembly in an old shell...

When GE bought them the archives were trashed! At least this is I heard in the past. RCA was then bought by Thompson and now by a Chinese company.
I would say that only the name survives.


polaraman

Makes me wonder if a new all-glass crt based on the later 21" tubes would be at all pracitical. It would be terribly expensive & I think that would kill it. Probably one of those things that would cost $10,000 per unit to cover the setup costs. And then it wouldn't be original.

The discussion of the adhesive is interesting. There are some amazing epoxies being made but I don't know how they work as far as heat. This may be a silly question, but I wonder if modern plastics have advanced to the point that a PLASTIC crt could be made? Thats kinda hard to imagine, I know. If it could somehow be done that would do quite a lot to reducing setup costs. That would have to be some tough plastic, though, wouldn't it?

Has anyone (with a little bit of CRT knowledge) thought to call Western Micronics and inquire about the 15gp22 rebuild process? Hell... it couldn't hurt to ask... and it would shed some light on whether or not these yahoos got all of their marbles!

There is always the possiblity (although unlikely) that they might actually have an "all-wise all-knowing big giant head" with the secret to this tube. :yes:

Some time ago I tried to contact them. Their phone had been disconnecteed.

Apparently, this would not be a good sign! :D :no:

[QUOTEWith time, the adhesive fails and air leaks through the seam. Restorers are experimenting with a new adhesive that has potential to keep air out for a long time, but the adhesive degrades at 500 degrees. Putting it in a 700-degree oven would not work. tube back to life.[/I][/QUOTE]

So why can't they bake the tube with the old, less-than-ideal sealant that can take the 700 degrees with no problem, then seal THAT seal over with the stuff that will not eventually leak?

Ed (AK user Colortel) told me about 25 years ago in this quest that a "frit" (unsure spelling???) was a piece of special glass that could be laid around the circumference of the tube at the seal that would transmutate into the glass and metal with heat during rebuilding.

Ed, if you read this, maybe you could refresh my memory, enlighten us and elaborate on that? I had kind of forgotten the conversation...

Charles

Response to 15GP22 Glass Frit comment
 

Charles,

Glass frit (you spelled it correctly) is essentially an extremely fine ground glass that flows into the microstructure of any adjacent surface (metal or glass). The frit is usually a lower temperature glass that has the ability to flow so that coalescence occurs during heating and bonds chemically to the adjacent surfaces. Temperatures as high as 500 degF (371 degC) not only helps the frit to flow but allows for rapid outgassing of the tube (mainly drive off water vapor - which sticks like hell to everything!). During manufacture high temperatures also help to anneal the glass structure. However, it could warp or buckle internal metal parts. As I recall, excess sealant exists during manufacture, so that during high temperature baking the sealant does not all pump out of the joint. However, reheating an old 15GP22 to extremely high temps would not be necessary since the original structure is already outgassed and frit seal created. If only the electron gun is replaced then outgassing may be accomplished at a lower temp, 392 degF (200 degC), within the limits even of any externally applied sealants, e.g. Varian 'Torr Seal'. The trick is to pump the tube for a long time!

'73
Tom Ryan

Did RCA or anybody else rebuild the 15GP22 back in the 50's or 60's? I think we are also reaching the day when even 21FJP22 and 21FBP22's and other delta tubes will not be able to be rebuilt. But there are a lot more of them around so it will be a long long time before they rare and they don't seem to have the vacume loss problem as the 15G does. The way things are going in the CRT field there is a small window left for the rebuild to happen or the sad fact is there won't much of a chance in a few years. I hope people are able to rebuild some of the tubes before the art is gone.
Steve

---

I e-mailed these folks after reading this thread. No reply yet, but the disconnected phone probably tells the story.

-Steve D.

I would think that bad 15GP22's in the 50's and 60's did not get rebuilt...since those sets were fairly rare there probably was enough new replacement tubes in RCA's stock to cover the need for replacements.

Plus the 21 inchers came out very soon after the 15s were introduced so by the time they needed rebuilt, it was probably a pretty good proposition for a customer to just dump the set and buy a lower priced brand new set with a MUCH bigger screen to boot.

Just out of curiousity...
 

What would a skid (10 or so) of 15gp22's NOS and PERFECT, in emissions and seal, be worth? (NO, I do NOT know where one is... sorry!!)

I am guessing, between 30 and 40k?

There may be such a cache of 15GP22's somewhere so it always pays for all of us to check out old TV shops, etc. before it's too late.
Whenever I'm traveling through little towns in Texas and NM and I come across any TV shops I ask if they have any old round tubes.

In case you haven't seen this story:

In the fifties, for the motion picture industry, the Hazeltine Company developed a system for viewing motion picture internegatives as a positive on a tricolor CRT. It is known as a color film analyzer, and the first one used a 15GP22.

The tube-type color film analyzers are now all gone -- the last all-tube survivor was a model made in the late fifties that ultimately used a 21FJP22. It was in India, still running in 1995.

About 1996, a Mexican company traded in an old analyzer that used the 15GP22. It was totally inoperable, but had been used up to around 1975.

The company from whom I got this information had an inventory of six NIB 15GP22's in stock until 1997, when management decided to toss all the tube-based analyzer inventory to make room. So along with all the other spare parts, there are six 15GP22's buried somewhere in a Los Angles landfill.

So yes, chances are good that somewhere there are undiscovered NIB 15G's.

The six spares in a landfill is a dreadful thought, but here is the flipside: Gary Hough (SoCal TV collector who passed away in '89) found two NOS 15GP22s back East. He bought them for a couple dollars, but found great disappointment in the fact that they had lost vacuum still new in the boxes. I had a similar experience with one that was in a working set, then the next time I fired it up it had failed to lost vacuum. The vacuum loss has nothing to do with how much if any time is logged through usage, so NOS is not applicable in the condition of a 15GP22.

I was told by someone that the manufacturing process involved some kind of procedure whereby the tube was cooked in a vacuum chamber without the front lens on it. Then the front lens and mask were assembled to the tube after cooking, still while inside a vacuum chamber. Zenith, not wanting to pay RCA, had to put visible screws in the mask of their 15GP22 clone because they couldn't figure out how to do it the same way RCA did... But the two known Zeniths still have full vacuum (KNOCK ON WOOD!!!) And one of the Zenith 43M20s with that tube survived being slammed repeatedly against a wall in the Northridge Earthquake... Enough to knock all the knobs off and break one of the wheels and still no glass-to-metal seal cracks! I was terrified to pull the back off that set for MONTHS after the quake, as I was convinced the tube had probably been wrecked. Finally I did and What a relief!

I met one of the 15GP22 assemblyline girls about 10 years ago, and she did recall some strange details. I hate to admit that I can't remember her name. She lives (or lived) in Westwood, CA near UCLA. I did a TV repair housecall for her, and the subject came up. I will think hard to see if I can get a clue and try to pull her up on the database. But even if I find her, I think she was at least 80 then. She was a WWII widow who had originally started at RCA during the war and kept the job on due to loss of husband in war, as I recall her telling.

Charles

Intriguing, are there details you can share?

Also, there are tales about RCA attempting to manufacture the 15GP22 in Puerto Rica and even Brazil. Did you glean whether the line she worked on was in Lancaster PA or perhaps there in CA?

---

It strikes me that this lady was from Back East. She remembered how they told her how "special" the tricolor tube was to be, and that she had to wear a space suit while working on it. That's about all the detail I can remember of the conversation, but I remember the type of TV it was that I repaired for her. So next time I visit the shop where that call came from, I will check their database to see if I can get her name & number from searching with the TV make and zipcode.

Charles

15GP22 Picture Brightness
 

Yes, that problem is characteristic of all vacuum technology, in general. But, the 15PG22 is a bit of an exception because it so happens that there's a lot of iron inside this tube. 15GP22 would have been an outgassing nightmare for RCA during tube manufacture. For a rebuild there should be much less water vapor to pump out - if the tube is not initially up to air and the rebuilder is clean and pre-baked all parts in dry nitrogen gas prior to installation into the old 15GP22. The more residual gas that's removed the less electron scattering occurs from the gun assembly when cathodes reach emission temperature. Less scattering effectively means somewhat lower emission temperatures are required to create a space charge around the cathode.

The real question to ask is:

How much residual gas pressure does it take to ruin a CRT, i.e. make the tube too dim to watch?

Well, it seems to me that the question depends on both electron gun efficiency and ultor voltage field intensity accelerating the electrons towards the phosphor screen. Bright pictures require more energetic electrons, and more of them, with high kinetic energy to produce more photons as a result of slamming into the phosphor dots at high velocity, hence giving us a brighter picture. As pressure inside a 15GP22 rises to 10-6 torr the rate at which residual oxygen molecules inside the tube strikes the phosphor dot plate rises to 3.6 x 10 to the 11th power (molecules per square centimeter). That's a hell of a lot of garbage in the way of the electron stream trying to hit a group of phosphor dots while being scanned at horizontal line rates! As the pressure inside the tube gets higher towards the 10-5 torr range the picture begins to noticeably lose brightness. Above 10-4 torr the picture starts to get really dim. At higher pressure less than 10-3 torr you can expect the tube to arc over inside - especially when the ultor sits above 20KV. Now, the getter should be turning white too! Also, you don't have to be at room air pressure to internally arc the tube. :nono:

Let’s put pressure into perspective: a new two-stage Edwards EM8 rotary pump (pretty common in most labs) after an oil change should reach 10-3 torr after 24 hours of continuous outgassing. To reach lower pressures requires either a diffusion pump or turbo molecular pump on the system to reach up to 10-6 torr to 10-7 torr range or maybe less (if a turbo pump). To get pressures lower takes either ION pumps or cryogenic pumps to reach the 10-9 range. Passive chemi-absorption techniques - hence the use of a 'getter' inside vacuum tubes can get things down to less 10-9 if the tubes seals are really good and keep it there.

I recall one interesting incident as a graduate student; I once cranked the voltage up on an electron detector in a mass spectrometer quite high one day - just to see what would happen. Even at less than 10-9 torr I was able to strip enough electrons off a detector plate once it reached 40KV to create a massive arc inside the vacuum chamber! Ka-Boom!!! The SS chamber didn't implode - just made a heck of a loud crack ...I jumped, probably 5 feet and ducker under a table. Yet, it always amazes me how modern large CRTs can work at 35KV or higher and not cause severe ionization inside a tube over time - pressures must be very low and kept very stable inside these tubes - a definite a challenge for old tubes! ...of yes, I forgot to mention the 40KV supply I used could deliver over 100 milliamps of current ...several orders of magnitude higher than a TV chassis!!!

Tom

I know this is a bit offtopic for this particular forum, but while we're on the subject of rebuilding tubes... I've heard of people that had set up shop for rebuilding black and white tubes in their houses - from what I'd heard, it seemed simple enough, a car battery and some nichrome wire to cut the neck off the tube, replace the gun and reseal the tube, and pump the air back out. But what kind of pressures are required for a black and white tube? And I'm only referring to all-glass tubes here - the metal cone kind would have the same sorts of sealing problems. But just how low does the pressure need to be in order to have a serviceable B&W tube?

Another thing that confused me about tube rebuilds (B&W tube) - did they replace the phosphors or only the gun assembly? I always thought that the phosphors in black and white tubes degraded with use too. Kinda like the screen burn on computer terminals where something would get burned into the tube if left long enough, and occasionally you'd see a really well used terminal with 24 rows of 80 fuzzy little boxes from where the characters lined up. Wouldn't a black and white tube eventually get the same sort of phosphor degredation over the entire tube? Or does the relatively rapidly changing content on television prevent this?

Basically, there's a lot that I don't know about picture tube rebuilds... Any insight?

-Ian

Sylvania rebuilt tubes were classified as "new gun" or "new gun and screen" so I think a lot of rebuilt tubes , B&W and color, were just equipped with new electron guns.

I really have not seen the kind of phosphor burn you get on old computer terminals on many TV crt's, except in sets from college student union lobbies or motel lobbies, etc. where the Weather channel or cable TV info channel played constantly for many years and burned text into the screen.

---

With regards to screen burn I've noticed a couple of things. Color CRTs don't seem to burn as "deeply" a mono CRTs. I've seen several mono CRTs that were used for either surveilance monitors or text displays where the burnt image is clearly readable when powered off. When powered on with a full white (or green, etc) raster the burnt areas weren't just dull, they were totally dark. This also happens with the old orange plasma displays used in some computers.

A while back here a load of ex-TAB 21" Orion TVs hit the market. The were used to display race information on teletext pages. When powered off several lines of text were clearly readable and several others visible as indistinct "shadows". When powered on it was only just noticable on a full white raster. Viewing a normal program you couldn't really see them unless you practically had your nose up against the screen looking for them. I sold several and never had any complaints or comebacks. As long as the customers got to see them running they were happy to take them, even after I explained the screen burn to them as showed them what they looked like when switched off. Of course it helped that I was selling them for half the price of a non-burnt set of the same size, and this was before the used TV market crashed. No-one would even ask about them when they were switched off, even when we dropped the price off the last couple of $20 to clear them out (this was back when a used 21" set would easily sell for $100, even without a remote).

I have a couple of Sony Trinitron studio monitors that were pulled from an edit suite. Both have time codes burnt into the bottom of the screen, but I don't really notice it. Otherwise they have beautiful bright sharp pictures with better color and geometry than many of the 1 year old sets that come through here. I like them enough to use them daily, one as a bedroom set and one in the workshop. Lately I've been watching more movies in the workshop than anywhere else (gotta test the DVD players before they go back out :) ).

I remember going to videogame/pinball auctions, and it was kind of fun looking at the old videogames with used color monitors where you could sometimes decipher what game the monitor had been in previously just by looking at the burn-in pattern : )

I haven't seen anywhere near as many burnt in arcade monitors as I would expect to see. Maybe around here the games were changed more often. The CRT rear projection monitors used in some big Sega and Namco cabinets are a different story though.

IRE paper discusses 15GP22 desgin and manufacture.
 

Good point. Actually there is an archive of information right under everyone's noses that was published in the proceedings of the IRE Vol. 42. Here it is:

Grimes, M.J, Grimm, A.C., Wilhelm, J.F. (1954). Improvement in the RCA Three-Beam Shadow-Mask Color Kinescope. Proceedings of the Institute of Radio Engineers. New York, N.Y., pp 315-326.

In this article Grimes, Grimm, and Wihelm describe the advances RCA has made in the design of the all glass color kinescope type 15GP22. Over what you ask? The original prototype was not the all glass tube everyone knows and fears can't be rebuilt today. The 15GP22 design originally started life as a modified 16AP4 metal shell. The 16AP4 was a glass-metal tube that found use in b&w sets that first appeared in the late 1940's. RCA simply added a color viewing screen assembly consisting of a shadow mask + phosphor dot plate and then added a tricolor electron gun. However, RCA quickly discovered a bunch of problems inherent to glass metal tubes adopted for color use. The authors interviewed many RCA engineers and assembly line workers to write their research paper. Maybe even the 80+ year old tech that Charles knows was interviewed. :D

The authors disclose a ton of information that led to changes in materials, assembly techniques and tips for factory production. For example, tube evacuation techniques, e.g. placing the mu metal shield outside the tube to reduce outgassing, improving the tension on the shadow mask by increasing heat dissipation 2.5 times to allow for increase beam intensities to produce brighter pictures, and changes to the blue phosphor to help reduce the classic "trailing edge" effect. The authors even reveal the composition of the shadow mask and the dichromating process used to improve thermal radiation and reduce buckling. Lots of other stuff is discussed too.
:tresbon:

But the real kicker for RCA in moving to an all glass tube design was that it couldn't build the glass metal color tubes fast enough. The authors reveal that RCA management wanted a new color tube to role off the assemble line every 2.5 minutes. The 15PG22 did improve upon the modified 16AP4 in three ways: improved tube performance, parts interchangeability, and decrease in tube cost. RCA eventually lost millions on this tube because the demands of mass production were never realized. The tube also had a high mortality rate in the field. Tubes were failing after assembly or losing their vacuum after delivery in new sets or just simply sitting in inventory. So, anyone who has a working 15GP22 tube needs to understand that it is essential to VacSeal seal all vacuum joints on this tube immediately to guard against the possibility of vacuum failure in the future. :nono:

Corning Glass Works originally manufactured the molds that made the face plate section and the cone (funnel) section of the 15GP22. Prior to assembly a metal flange is sealed onto each part separately. This was done by RCA in a refractory oven using similar techniques for b&w tubes. Glass frit is used to bond the metal flange surface to the glass at extremely high temperatures. NOTE: at this point you are only sealing metal to separate glass parts. They are not yet joined together to form a CRT. The second major assembly step was the use of a very complex jig called an assembly lighthouse. The lighthouse was used to set the mechanical position of the phosphor dot plate to the shadow mask. A special jig is also needed to secure the shadow mask to the cone section flange collet assembly at the correct tension. The phosphor dot plate was then attached to a collet assembly on the face plate section. Once the alignment was completed in the lighthouse jig the flanges of the two halves were then heli-arc welded together. :smoke:

From my experience, a tri-color electron gun assembly may be substituted into a 15GP22 as long as it meets one very critical requirement: the electron beams from the gun must focus at the plane of the phosphor dot plate exactly as simulated using the small light source in the alignment lighthouse during manufacture of the original tube. This would first require precise mechanical alignment during a rebuild so that the replacement gun is attached at the correct position, and establishing the correct electron field density to focus the electron beam during operation. :scratch2:

Finally, I would like to share with everyone what the authors had to say about baking the assembled 15GP22 after final assembly. This was especially critical during outgassing and evacuation. The authors pointed out that the manufacturing of b&w tubes used higher temperatures which were found not suitable for manufacturing color tubes. Why? Slow thermal cycling is required to prevent the shadow mask from bucking. The face plate must be also shielded to prevent it from cracking due to non-uniform thermal shock. Most importantly baking temperatures were tipped off at 300 degC. :yes:

See, when you stop and think about it the 15GP22 resulted because RCA engineers failed in making the "old design" 16AP4 glass-metal color tube work cost effectively. Grimes, Grimm, and Wihelm provide a wealth of information in their article about 15GP22 construction and assembly. The article also reveals what later become known as common sense techniques for color tube production: success depends on using moderate temperatures, slow thermal cycling and long tube evaluation time. :thmbsp:

Tom Ryan

<<<I always thought that the phosphors in black and white tubes degraded with use too. >>>

Since back then there were no news channel crawls, logos, videogames etc. it would have been evenly worn as long as the set was working right. I can see if something went wrong like you lost vertical for hours while nobody was tending to the set and it burned a horizontal line into the phosphors. Or the ion trap got messed up and it ended up causing a burn spot in the center.
Otherwise the big % of the cause of dimness in an old tube is probably the gun's fault or because of outgassing. Plus doing the guns is hard enough, I can imagine the pain it might be to remove/replace the phosphors. How was that done anyway? Or did they even do it?

On the side-note of color tubes with burn, you don't _normally_ see modern color computer monitors with burn, until you start working with junked equipment from state offices, hospitals, etc. Server consoles are pretty bad too - I've seen monitors that very clearly show "Windows NT Server" or "Windows 2000 Server", burned sharply into the screen. Once I even saw one with Windows 2000 burn _and_ NT Server burn. They must have upgraded :). Or offices that have their company logo as the desktop wallpaper, those typically get etched into the screen pretty good. You also see login screens burned into the tube quite a lot. Occasionally, you can see the whole desktop, complete with icons and readable text - even the icon headings of "Microsoft Word" are readable. And that little MS word 'toolbar' thing.

Arcade machines, especially the old ones, get burn in attract mode, or just from use. You can always tell a monitor from a Pac-Man or Ms. Pac-Man machine, the maze is clear as day. One time I even saw a TV set in a thrift store with Atari 2600 Pac-Man burn in it.

I guess that under _normal_ use, the phosphors in a TV shouldn't degrade too much, so rebuilding the tube would be only a replacement gun. It seemed like I read somewhere that the really old B&W tubes suffered worse phosphor degredation - until they switched to an aluminized screen or something like that. But the tube rebuilders probably would look very closely at the duds as they came in, and only rebuild the ones that were going to work well when rebuilt. But I am really curious, if they did replace phosphors in the tube at some point - how did they do it? I can just see someone trying to scrub the old phosphor off the tube with a long brush or something, like cleaning out a bottle.

-Ian

<<<It seemed like I read somewhere that the really old B&W tubes suffered worse phosphor degredation - until they switched to an aluminized screen or something like that.>>

Very early tubes didn't trap the ions so would get an ion burn spot eventually, later they added angled guns + ion trap magnets to suck up the ions, then came phosphors coated with very thin film of aluminum which stopped the ions and eliminated need for traps.

Tom, the 15GP22 is different from other guns in two ways: First, it uses high voltage (around 5 kv) focus, and second, it has electrostatic convergence, with an additional element in each gun. There have been a couple of attempts to use 21AXP22 guns in 15G rebuilds, the assumption being that the screen size is so small that convergence would be adequate without dynamic convergence. I've never seen the results, though.

Black and white tube rebuilds were done with new phosphor (more expensive), or re-using the existing phosphor. This is still true today - Clinton Electronics is attempting a rebuild of the prewar 12AP4, a tube that didn't use an ion trap and therefore had ion burns. They remove the old phosphor and re-coat the tube:

http://www.earlytelevision.org/prewar_crt_rebuild.html