Yeah, see, I know this guy, and he’s got this big problem. He’ll make this clear neon filled tube - and I swear this is true, I’ve seen it - it looks great on the burn in table. Beautiful red from end to end as soon as he turns it on. But here’s my friend’s problem: the tube runs for 30 minutes or so and it starts to turn pink in the middle of the tube! Just sitting there changing colors! It’s about to drive him crazy!

Well, it’s not just one guy that has seen this problem - it’s a bunch of people. And the first shop that I know of who dealt with this problem in secrecy was ours. Not that others hadn’t had this problem earlier, it’s just that they weren’t talking about it either. No one wants their customers to find out that they have some problem that they don’t understand and can’t solve (except by either remaking or repumping the pieces, which has put you so far behind schedule that you don’t have time to talk to anyone; and I mean really behind because you’ve already torn your manifold apart and cleaned it, changed the pump oils, and replaced the gas bottles MORE THAN ONCE, and by this time you’re just so damn frustrated that no one wants to talk to you anyway).

And the ironic part is that this problem is only seen by those people who’s red tubes turn on perfect red. Those people who have spent the money on great manifolds and gauges and invested time in learning about the bombarding process; not your competition down the street who makes those ransom note looking window signs where all the tubes snake. But this makes sense: you couldn’t possibly see a slight color shift in a red tube that turns on pale blue and snaking and takes a full day to become red. Not that you’re seeing the humor in this when it’s happening to you.

The problem is this:

Bombard and fill a clear tube with neon; on the burn in table the tube turns on red end to end with no snaking / side discharge / or any other problems. After the tube has burned in (using 30mA) for 30 minutes or so there is a color shift to pink that starts in the middle of the tube. In the best case it will recover within 30 minutes without becoming hot in the pink section. In the worst case the pink will spread over the entire length of the tube and it will become very hot to the touch. It may recover in a couple of days (but usually by this time we’ve usually thrown the thing away). I have heard of the problem not manifesting itself until much longer - perhaps days of running.

Over the past few years we have had this problem off and on. The "solution" (that would be after the teardown/clean paranoia stage) would be to switch to another manufacturer’s glass. This is the obvious test to isolate the problem as a glass problem rather than a manifold problem and we might not have the problem for a year or so. In talking to the companies that make the glass tubing, phosphor coat it, make bombarding equipment, and regular old neon people, I have not heard a single simple solution to the problem. I believe that this is because there may not be a single cause of the problem. Just like good neon comes from doing many small things right, troublesome neon usually is not caused by just one thing, but several problems adding up to put you over the line from "looks good" to "oops, big problem".

Two of the companies that have been very helpful in getting information out on this problem are Eurocom and Osram Sylvania. Although they are writing about different causes of problems in glass, that does not mean that both are not correct - that there is more than one factor working here to make the tubes bad. I’ll start here with the (1994) Eurocom technical bulletin because they address the problem straight from a tubebenders perspective:

Re: Clear Red Neon Gas Filled Tube Color Shift After Installation

Dear valued client:

In reference to our many conversations in recent times concerning the color shift that your clients are seeing in recently completed clear red tubes, I would like to clarify the cause and suggest the remedy. This scenario has been witnessed periodically in North America as well as countries in Europe for approximately three years. It occurs with all types of electrodes, tubing and pumping systems in both large and small companies.

In recent times, we have been called on to explain this phenomenon by firms around the country and Canada, many of whom are not our clients. Physicists at our corporate location in Aachen, Germany have been studying this along with the Schott engineers for some time.

Although the findings are incomplete, some answers have surfaced:
It has nothing to do with electrodes, vacuum pump or manifold used.

The color shift is caused by the glass releasing occluded gases in part due to some changes in the structure of the glass that have been made in recent years.

Some batches are worse than others and one should run a simple test when opening a new box of clear tubing without phosphor coating in order to establish the quality of the glass.

In most cases the tube color will return to red after some time, possibly in a day or two, however, one should avoid the occurrence whenever possible.

The technician should make and process a 15 mm clear tube with the 80 R 65 electrodes, bringing the tube to 350 degrees Celsius to thoroughly bake out the silicates.

The tube should then be allowed to burn-in on 30 mA current for a minimum of six to eight hours for observation, this problem will in most cases be eliminated.

Note that the tubes come off the manifold with a bright red color and if a change is to occur, it will begin in 20 to 30 minutes, however in some cases it requires a few days, beginning with the center of the tube. When a particular batch of tubing is especially bad, return the box to your sign supply house and ask for a replacement. Under no circumstances use 60 mA current for either burn-in or installation of the sign as this higher current will tend to cause release of these undesirable elements within the tube.

This might seem extreme to someone who has not been troubled with this challenge before, but it is necessary to deal with it until the cause is completely identified and corrected. Our people are committed to this project and plan to offer a processed clear glass free of this problem at some time in the future. However, today what we have outlined above is one way we can deal with what is presently on the market. Your client must know that this should not reflect on a negative way on your product quality as it is witnessed in some of the finest companies working with the best equipment available.

A couple of notes:

1) the temperature that is suggested (350 C) is much hotter than most of us get glass. I have tended to see the problem lessened by heating more. There are others who I respect who have found the condition worsened by high temps (much like uncoated ruby). Maybe this simply means that there is more than one source of similar looking problems. Maybe I don’t know.

2) my British friends say that we Americans should be damn happy that we don’t have this problem as bad as they do using soda clear. I understand their point, but it still is a pain dealing with the problems we have. I know of one person who believes that he is having this problem with lead free; which doesn’t seem likely, but possible. Again, similar looking problems etc., etc.

Osram - Sylvania has been looking at another (different ?) source of problems in glass tubing. Note here that neither I nor they are saying that this is the sole or partial cause of these shifting to pink problems, only that this weathering phenomena causes some problems in tubing. There are many variables when it comes to making good neon. It is difficult to isolate some variables without affecting other variables. But there are many similarities to what is understood here and what I’ve seen in my shop.

Read on:

Weathering is the product of sulfur dioxide and carbon dioxide combining with the soda in the glass and water in the air to form crystals of sodium sulfate and sodium carbonate. This process begins as soon as the glass is cooled down after the initial manufacture.

The humidity conditions at the time of manufacture play an important role in the speed of the weathering crystal formation, and the humidity and ambient temperature during the storage and shipping of the glass play a secondary role after the initial forming.

The crystals begin singular columns, like grass growing out of the surface of the glass. As the columns get longer, they eventually fall over, and grow along the surface of the glass. Dr Tichane told me that the crystals proceed through 23 different formations. As the soda on the surface of the glass is used up, additional soda will migrate to the surface to replace it. The ultimate is a crystal formation which is a series of needle like crystals that wander about the glass surface resembling a Chinese dragon.

Generally we find that the weathering forms faster on the interior surface of bulbs and tubing than it does on the exterior surface. This is because during the time that the glass is very hot, some sodium volatilizes from the glass. The air inside of the tubing or bulbs is trapped, so when the glass cools enough for the sodium to condense, it condenses on the inside surface. The air on the outside of the tubing or bulb is not trapped, and so may cool elsewhere and leave the sodium to condense on other surfaces. Thus the surface on the interior of the tubing or bulbs is soda rich, while the exterior surface has the same level as the bulk of the glass.

Both sodium sulfate and sodium carbonate are water soluble. Sometimes when the glass is being stored where there are warm days and cold nights, the air in the tubing is expelled during the day as the glass heats up. At night, the air and the tubing cools sucking in new air which has a high humidity. Later at night, condensation forms on the inside of the tubing, which dissolves the sodium carbonates and sulfates. The next day, the condensation evaporates leaving the sodium carbonates and sulfates as water marks. If the condensate is heavy enough, it will run down to the inside bottom of the tube, thus leaving a string of watermarks on the inside of the tubing.

Many formations of weathering can be washed off with just water. Tubing that has been washed will weather again, but at a very different rate. Of course, when glass is washed, care must be given that the glass is well rinsed, or when the glass dries, there will be water marks. Sometimes water marks concentrate the products of weathering so that the water marks become a bigger problem than the original weathering. In early phases of weathering, the grass-like stage, can be dissolved with just condensation.

Weathering can be decomposed with heat. In the early to middle stages, if the glass is heated to above 475 degrees C, the sodium carbonates and sulfates breakdown in to water, sulfur, or carbon dioxide, and the sodium goes back into solution in the glass. This is one of the goals of bombarding neon lamps during the exhaust process. Phosphor coated neon tubing, while it has been washed and heated during the baking of phosphor, has probably weathered again to some degree, so the bombarding of the lamp is required to remove these products. Clear neon tubing probably has never been washed or baked out and so the weathering products have accumulated from the time the tubing was formed. Tests have shown that clear neon tubing lamps require a little extra heat to clean them up so they will not discolor with age.

Weathering in the very advanced stages may require longer times and additional heat to achieve complete decomposition of the weathering products. If the glass is heated too hot too quickly, the products of weathering may get burned into the glass surface just as heavy water marks can become burned into the surface. When this happens, the surface becomes pitted and these pits act as sources for fractures and thus the tubing becomes brittle.

Unfortunately, there are so many variables that effect the rate of weathering, the glass composition, humidity and temperature at the time of forming, humidity and temperature as the glass ages, the amount of time at various humidity and temperatures, the cycling, the level of SO / 2 and CO /2 in the atmosphere, the level of soda at the surface, etc.; that it is impossible to make very definitive statements regarding the rate at which weathering will proceed or the maximum age of the glassware before weathering becomes a problem.

Because of the many levels and forms of weathering, it also becomes difficult to state absolute cleaning processes that will work 100% of the time. Each situation must be evaluated on it’s own. Fortunately the extreme conditions mentioned in the above paragraph where the removal of weathering becomes a problem are caused by unusual conditions, such as extremely high humidity, extreme levels of SO /2, or extremely old glassware (several years).

For the most part, glass users have found methods of dealing with weathering that are effective. For example, given the composition of neon tubing, normal turn over of inventories, and the normal atmospheric conditions that these inventories are subjected to, the bombarding procedures recommended in the neon industry are effective in cleaning up over 99% of these conditions. However if someone stored tubing for several years in sauna like conditions, he would have problems.

We have tried to pass on what we can. We do not mean to make the weathering problem sound worse than it really is because we are dealing with the exceptions than the rule.

A couple more points:

1) during the first year that we had this problem I managed to get some glass that had gone through the cleaning process that coated tubing gets, but it was left uncoated. I could use the "regular" glass and have these color shifting problems and then switch to making identical pieces out of the "super clean" glass and the problem would not occur. This was wonderful: it gave us a benchmark glass to test all others against. We could eliminate the manifold / gas as a possible source of the problem. We set this special box of glass aside.

We switched glass manufacturers and the problem went away. It didn’t show back up until two years later. No problem we thought: we can just pull out our box of special glass to confirm that it’s a glass problem and get on with our carefree bending. Perhaps we were gloating about our incredible cleverness some while making our benchmark test pieces. As you may guess, the test pieces shifted color also. We had not known this stuff about sodium leeching out over time (and ruining our "benchmark"). And yes we suckered into cleaning the manifold, changing pump oil, and throwing away perfectly good slightly used gas bottles. Again. Only when we went back to the first glass manufacturer did the problem go away.

2) the high temperature (475 degree C) mentioned as the temperature necessary to decompose weathering can is not the only way that weathering can be decomposed; lower temperatures for longer periods of time are also effective.

So if we can’t really pinpoint a specific cause of the problem, how can we pinpoint a specific solution?

We can’t. Or at least I can’t. But here are some things that you can do that will help:

1) decide how bad the problem really is.

If the color shift is VERY slight and it burns in within an hour or so, I really wouldn’t worry about it. Yes your electrodes are having to getter this stuff and it is probably shortening the life of the tube, but we’ve got some six year old stuff that had these problems that looks great today. Having said that, anytime the color shift was accompanied by the tube running hot we trashed it and started over. This leaves a lot of choices between these two extremes. It’s your call.

2) change glass.

If the problem is severe you might consider throwing out the glass you have, cutting your losses and getting your glass from another distributor / source. The time it takes to re-electrode a big job (if that is the solution you need) isn’t worth the cost of a box of glass.

3) repump the tubes.

If it’s the scrubbing of the arc stream or just time that liberates this trash, the burn in (or in this case the anti burn in) time will free the trash from the wall of the tube (dumping it into the arc stream). Unfortunately there may be more trash for the electrodes to getter than is reasonable.

Everyone that I have spoken to who has had this problem has had excellent results by repumping the unit. (Another reason to use quality electrodes). Very time consuming but this can save the tube. If you’re having this problem, don’t tip off short until you’re sure the tube isn’t going to turn pink.

4) I believe that preheating does great things for tubes in almost all situations. Except for this one. Not enough heat for not enough time is my best guess as to why.

5) tweak everything else.

Don’t add to your problems by having less than your best performance out of your manifold.

6) be consistent.

Make your pumping process repeatable. If there is a problem, you can experiment until you find a solution for that particular problem that you are having. I have seen this problem dramatically reduced by a simple15 degree temperature change. Gauges let you establish a baseline, understand what effect small known changes make, and apply the best procedure for your situation.

Many thanks to the people at Eurocom and Osram - Sylvania for all their help.

Morgan Crook is a mechanical engineer, tube bender, and president of Neon Design, a wholesale, retail and consulting firm in Columbia, S.C.