Feel’n the Heat.

I am a geek. Anyone reading many of my posts already knows this. I am also a collector (my wife says “hoarder”….listen to her, she is very wise) of vintage technology and neon sign production tools in particular. But again, if you have been here before, you know this. Not content to simply collect my various antiques, I like to use them…at least a little…..certainly enough to enjoy them or otherwise have first hand knowledge of what they were like to use day in and day out. Such items are not only comforting to me (some deep rooted quirk of my childhood perhaps) but plain and simply fun for me. That ’47 DeSoto, for example…..it is my daily driver. (at least during the snow free months) So it should come as no big surprise that when I learn of some odd example of some piece of equipment I will most likely seek one out to add to the collection, use, or at least try out….some such items gain a place of prominence in my shop to be used regularly, others relegated to the multitude of shelves awaiting my further curiosity or that of a visitor. And some, in that process, get a little publicity right here on this website enabling those who are curious to have a vicarious visit as it were.

This specific object is an old (likely 1920’s or 30’s) Thordarson temperature gauge. It was designed for use in neon shops to measure the glass temperature while processing the tubing unit. Those who are familiar with neon or who have read my “Neon 101” article will understand. In short, when evacuating the tube it is necessary to heat the unit up in order to liberate impurities from the glass and electrodes so that the vacuum pumps can carry them away. But HOW HOT? …and HOW would you know?…….you sure aren’t gonna touch it while doing this….the temperatures needed are on the order of 450-550F and the bombarder current used to obtain this heating is lethal.

Many benders, including myself, were taught to place some news paper strips on the tube. When they charred, it was “hot enough” and you could stop and let the pumps do their work until ready to fill. This is not really a great method for reasons that I shall explain later. In spite of this, it is not entirely uncommon, even these days. The other two common methods used are the Thermocouple gauge and the Infrared gauges. One additional way was with the use of thermal crayons that change color at a certain temperature, but I never particularly liked them in part due to the mess and they seemed to be only marginally better than the newspaper method.

A few years ago, however, I saw a photograph of an old mechanical gauge made for this purpose. Likely of 1920’s or 1930’s vintage, it consisted of a little footed bracket and a bimetal strip wound along a shaft. This would deflect the pointer across its scale. The unit was marked in tens of degrees F and had tick marks placed for target temperatures of standard lead/soda glass, pyrex, and the cool down fill and seal-off points. Thordarson, the famous maker of transformers, produced this during their years of manufacturing neon shop supplies. But. I had never actually seen an example of this device, nor had most of whom I had spoken with, and many had never known such a gauge was made at all. Naturally this resulted in a quest…because that’s what I do.

After quite a long time, I spoke with a gentleman who had collected sign shop related items. (see, I knew I wasn’t the only one!) After some discussion, and more waiting to get everything sorted out, I was able to purchase an example of this gauge from him. It arrived while I was away on a business trip and Elizabeth carefully placed it on my bench awaiting my return. After getting back and looking it over, I found it to be in very good shape and could thus proceed to try it out. I had been wanting to try one of these since learning of their existence. Not so much to use day in and day out, but I wanted to see how its performance compared to the gauges and methods I was accustomed to using.

My go-to method these days is a thermocouple gauge. I have an old Haydu Bros unit from the 40’s and a slightly newer one from Scientific Glass Apparatus. The latter one is in a convenient heavy base and gets used regularly. These operate because two different metals, joined together, will generate a small current when that junction is heated. Such devices are present in many gas furnaces, and were even used to generate power for spacecraft via the heat of radioactive decay. My example, however, may not really be spot on accurate….although it seems close. Really, the true numerical value reported is not terribly important so long as it is a repeatable measurement and that it is interpreted correctly. Afterall, you can call boiling water any number of degrees you wish, so long as you always use that value and scale and recognize what the chosen number actually means. In the case of a neon plant, this becomes a bit of a hybridization between the hard science and the intuitive art that comes from familiarity of the processing system being used…..old-timer knowledge as it were.

The other gauge I have is an IR thermometer. I do not use it in my neon work as it is intended for other mechanical uses and is not really configured to this purpose. But since I had it, I wanted to use it in this test as well, just for comparison. IR thermometers function by measuring the infrared emissions from a heated object. Their accuracy is influenced by the emissivity of the specific surface being scanned. Some materials emit more IR at a given temperature than others. My Raytek gauge has the ability to be calibrated for the albedo of a specific material if it is known. Unfortunately, I do not have any such information for the glass and so I just left it at its default setting for this test.

All of the above information should give a bit of a clue why the old newspaper method is not recommended. Both accuracy and repeatability are suspect. Newspaper chars. But at what temperature?…well, certainly by 451 F ….Ray Bradbury said so! …but it begins to do so before hitting that mark, and continues to increase in charred appearance afterward…..humidity will affect this value. So will the weight of the paper and what type of inks were printed upon it prior to being cut out of the newspaper. For these reasons, it is really not a good gauge of temperature except in the most course of sense…….and for those who just like to smell a little smoke as they come close to burning something….which may be its most redeeming quality for this application.

So, this morning I took a straight 12mm tube that was handy, put electrodes on it, and sealed it to the pump system. Once connected, I fitted the usual thermocouple gauge, a couple of strips of newspaper, and the Thordarson gauge. It was a little hard to get it placed well, but stayed put once situated. I kept the IR gauge handy as well.

I processed this unit as I would if I intended to fill and use it and made note of the various gauge readings and points of where the newspaper charred. I then let the tube cool off, vented to atmosphere, and did it again. I repeated the process three times in order to gain some insight into the repeatability aspect of the Thordarson gauge.

As the initial heating took place, the Thordarson gauge actually reacted much faster than my tried and true thermocouple meter. This was a little bit of a surprise because the construction of the gauge keeps the bimetal coil about 1/16″ above the glass rather than in contact with it…I had suspected that the air cushion would have slowed the measurement rate. This theory was promptly disproven. In addition, at lower temperatures, the thermocouple and the mechanical gauge were in very close agreement.

The IR gauge could not be used while the tube was lit, as the extra light emitted interfered with the measurement by over a hundred degrees. It also tended to be all over the place and was overly sensitive to where on the glass it was pointed. This is why I do not use it for this purpose.

The thermocouple gauge, being my standard, behaved as it always has. Its response time was slow, but the measurements consistent and usually within expectations.

Not surprisingly, the newsprint charred well below the target temperatures needed and began to smoke a lot. Ah, good memories of my youth. But not always good tubes if you depend on this method.

The Thordarson gauge tracked faster than the thermocouple gauge but stayed within 25-50 degrees of it for most stages. This was a bit better than anticipated and the error between the two was almost nonexistent up to about 300-400F, after this, however, the Thordarson overshot rapidly and gave an obvious false reading that greatly exceeded the thermocouple or even the IR gauge values.

I found when repeating the test that the results were pretty much the same, within a much tighter range that I originally expected. The gauge seemed repeatable within 10-35 degrees each time…..maybe better if I paid closer attention……it is hard to do that while trying to take notes, run the process, and take photos all at once.

As for the increasing error with increased temperatures, getting back to that “old-timer knowledge” ..IF you knew this, and IF you paid attention to when it begins to rapidly overshoot and interpret it correctly, there is no reason that you could not learn to use this gauge and get suitable results.

Perhaps I shall continue to use this gauge for a bit and see how the resulting tube units perform over time. Even if I don’t, however, this has been a fun little experiment to follow up the quest to obtain what is likely one of the rarer pieces of vintage neon shop hardware and this gauge will enjoy a prominent place in my shop collection.

I