Identifying and Dating Vintage Neon Transformers

NOTE:  This post is a little different in that it is one I expect, actually hope, to be able to update and revise as new information becomes available.  If this proves to be the case, I will be sure to specify the date that I made any additions or corrections to it and keep things such that it will be possible to keep track of what was learned along the way.

On one of the neon groups that I am a member of, we have had several of the posts along the lines of “what year is this transformer from?” when someone has an old sign they are servicing or restoring.  The value of this information goes a little further than simply answering the idle curiosity….although curiosity is more than enough reason for me to pursue the topic.  In a more practical sense, such information gives a good indication of how old the sign really is or at least when it had last required a new transformer during a servicing event.  You will often be amazed at how long this stuff lasts!

When dealing with many vintage items, it is often helpful to look closely at construction and design techniques as these are a good indicator of the general age.  This cannot be confined to the outwardly visible parts, but must be carried right down to individual components and even the wire types used.  (see “Look at These Bits of Wire”)  In the case of antique radios, for example, it could be possible to replicate a wood case….but much harder to replicate internal components.  Such information can therefore be helpful in assessing the authenticity of an item, or at least when it was last serviced or restored.  In a world where fake items are prevalent, this can be of great value to the collector or historian.  For now, I will confine this post to the subject of neon sign transformers.  After answering a few group posts on the topic and looking over older transformers that I have chosen to keep over the years (more on why I keep some will follow) I have decided to try to come up with a little bit of a quick reference for others to try to get into the ballpark even if an exact date of production is not to be found.

At some point, certainly by the 1970’s and 1980’s, most of the remaining manufacturers of neon sign transformers began using a simple two digit month/two digit year as the serial number.  Some separated it with a “/” and sometimes the numbers were just run together such as “0486” or “8604”which would denote an April of 1986 production.  These are readily sorted out and so I will leave those alone.  The older units, however, are another matter.

From the beginning of mass produced neon transformers in the 1920’s on into the 1960’s, the model numbering and serial numbering schemes seemed to vary greatly by manufacturer and may very well have been changed from one period to the next within a single company’s production range.  My sample to work from is by no means complete, and I welcome any specific photographs and information that others with experience on this may offer and I will try to add revisions to this post if new information arises.

Many of the internal construction features of a transformer are common across the board– laminated iron core, fishtape, wire windings, and potting compound, etc, are all very straight forward and not wide in variations from one make to the next…However, the outer casing and insulators were sometimes treated as if it were the body of a fine automobile.  A look at some of General Electric’s transformers of the 1930’s and 40’s is all one would need to support this contention.

The golden age of industrial design was not limited to the outward appearance of a finished product.  The beautiful streamlined art-deco style was even applied to some internal components.  This level of attention paid to otherwise basic parts that would never be seen except by the builder or service technician is, to me at least, a mark of high quality and was no doubt one of the first things to be cast aside in the modern era where cost is the primary driver of nearly all decisions.  How painful it is to those who believe in their artistry, even on mundane items, to have specifications dictated by the accounting department.

GE was only one maker, others such as Sola and Thordarson made these as well.  Generally speaking, these case styles were done away with sometime in the mid to late 1950’s, although some rounded corners remained for many years even if they did not possess the full on streamlined look of their predecessors.  This is not to say that every art-deco cased one is of this era or that every straight slab sided cased one is newer–as there are many possible exceptions….but….as a rule of thumb….if your transformer looks like it wants to fly through space on its own, then it is more than likely of this period.  These streamliners really are functional works of art…I keep them when I find them….at some point I hope to make some neon artwork that allows them to be visible rather than hidden away in a sign can, as they are worthy of being viewed in their own right.

Some manufacturers used a long model number and others used something simple that denoted the output specs.  Most are familiar with seeing something like “12060” or “1260”  either of which would most commonly be a 12000 volt 60 milliamp model.  Sometimes it was “930” or “9030” for a 9000v 30ma unit.  You get the idea, and in all cases the tag would have these specs on it someplace else anyway regardless of the model numbering scheme used.

Unless a separate specific date code block was shown on the tag (and if it is, lucky you as you have no need to guess) then it was very common to use the serial number as the date.  This is where things become tricky on the older units…I am uncertain that any two makers ever used the same scheme for date codes….I have seen letter/number combinations…number/number…..and some letter/letter codes….some with 2 digits for the year and some with only 1 digit for the year.  This at least lets you get it right within a decade…and then case styling and knowledge of what it was installed in can be of help in narrowing it down to which decade it was made.  Unless, of course, no serial number is given at all.  Yes, there are models out there with no apparent numbering or lettering that can readily be translated to a date….for these, you will have to look closely at case style, insulator bushing type  (is it a simple straight design or tapered in some way?), and if you have the information available-what it was originally installed to.  As a rule of thumb, screwed together cases and tags with tapered insulators are older and rivets with simpler designs are newer.

During this era, General Electric appears to have used a two-letter coding system that I have yet to decipher.  I have seen examples showing EW, PG, RE, WD, and a few others….all of which were on 51G series transformers–some with streamlined cases and some without.  I can tell, from looking at the catalog numbering, that 51G is obviously a neon transformer, the following numbers have to do with output……and….IF a second number 1 is added, it appears to denote a high power factor corrected unit.  For example, 51G1 is the 15000v, 30ma unit…..but I have also seen a 51G11 that is also 15000/30 and looks identical but identifies it being power factor corrected.  A few others I have found also appear to follow this convention.

Some of the letter/number schemes work around a letter for each month….likely starting with A for 1 (January) and proceeding onward, possibly skipping a letter or two if necessary (I suspect the letter “I” might get skipped since if the tag was dirty or dinged it might get misinterpreted as a “1”) ….Others, however, appear to use letters of the month or perhaps a letter code for the week or bi-weekly period of the year……the latter fits well with the 26 two week periods of the year……and so you may see an Acme (Hey, even Wiley Coyote wanted neon and who else could he have possibly ordered it from other than Acme!??!) for example….showing a P51 code.  Likely 1951….but what did the “P” signify?….perhaps it was the 16th bi-week of that year (around the beginning of July?)…..this is the sort of question I hope to one day be able to answer with a degree of certainty, but for now getting the year correct is helpful to many even if the week or month is still a guess.

Another oddity I have encountered but have only seen once is an old Jefferson transformer, a fairly small one at only 5000 volts, that has a separate metal tag on it….this tag is a footage chart!  I am still trying to determine what years this was done and if it was restricted to any specific series or models.  I do know from looking at a lot of them, that Jefferson (later a division of Magnetek) did not use date codes for their old stuff.  I guess they just figured their products would last so long no one would ever need to know…..and….I must confess, many of their old ones I have encountered do in fact still work!

By now you may have noticed that I have not said anything about patent numbers.  Patent numbers are helpful, but in my opinion only up to a point–a patent is valid for 17 yrs and so does not really narrow things down much other than being able to say that it was made after whatever the most recent patent date is.  A more helpful scenario would be to find two identical units, one with a patent pending statement and the other with the patent number…then you could likely state that the pending model was made between the filing date and the time the number was issued.  That, too, could be a couple of years or more.  …and it is an unlikely situation to run into at this stage of the game.

For the most part, when tracking these sorts of things down, old catalogs/brochures and trade publications with ads are helpful….but often the biggest help we likely will encounter today is simply seeing more of the old units themselves.  Companies have come and gone, and so have their records for such information….unlike, say, the Hasselblad ledger books that detail each camera and who bought it back in the 1950’s and 1960’s, information on neon transformers is far less specific and is mostly going to be determined via anecdotal and perhaps even circumstantial evidence.  A larger sampling, complete with as much ancillary information as can be found, will make cracking the codes used a little more likely.  For now, I hope this has been helpful in pointing out a few basic things to look for when you find yourself asking how old one of these is.

 

***update #1—-new information added 11/21/16***

While traveling this winter we have been spending some time at my parent’s house.  My father collects vintage technical books and catalogs:  one of the interests that I obviously inherited.  While perusing his bookshelf I found a Westinghouse Electrical Supplies catalog.  This catalog from 1936, like many of the era, is worthy of a website in and of itself.  For purposes of this discussion, I am adding some photos of pages detailing their offerings of Neon transformers at the time.

Of interesting note is the discussion of the balanced secondaries that would offer protection in the event of one side being shorted.  In addition, the units designed with built-in electrode sockets are of interest not only from the skeleton sign standpoint, but also because they mention their use for free standing neon pieces.  This may well be one of the first incarnations of the familiar “Neo-base” upon which some sculptural works are fitted today.

The following photos are of the other luminous tube transformers depicted in the 1936 catalog.  In addition to the images and descriptions, the catalog numbers may be of value to others as they will help with dating similar units.  It can be assumed that these were not “new for 1936” and that they were also produced for a time after that…therefore if you encounter a Westinghouse unit with these numbers it is likely from 1936, give or take a few years.  (likely by 1942 production ceased for the duration of the war and although pure speculation on my part, the postwar offerings probably had different designations)

 

I hope you find this additional information interesting and valuable and I look forward to adding more when it becomes available to me.

 

****UPDATE #2—-New information added 4/29/18****

I recently came into possession of a Jefferson transformer catalog for their luminous tube offerings.  As previously mentioned, Jefferson was one of the top notch builders of neon sign transformers of the day and many of their vintage units I encounter are found to still be operational.  The Bulletin number 402-LT carries a 7-46 publication mark on the back cover and this example has a “Received 7/23/47” marking stamped inside so I am calling this a 1946/47 catalog.  Chances are this was the first “new” catalog they had in the postwar period as civilian products of this sort were unavailable during the 1942-45 period due to wartime production needs.

The catalog goes into great detail on a variety of their offerings and includes design and construction information as well as citing patent numbers where applicable.  In addition to the range of sign transformers offered there is also detail on their bombarding transformers of 10 and 15kva capacity and the slide type control chokes for them.

One interesting clue for dating these units is the statement that they “Now” provided a separate grounding lug on the transformer case of models 721, 722, 724, and 725 to comply with UL code requirements.  One can safely assume then, that this had not been a required feature for that class of transformer prior to WWII…so if you find a similar unit lacking this, it can be safely guessed that it was made before 1942.  This may not be common to other manufacturers or to types other than those that are referenced as such in the catalog so exercise some caution before using Jefferson’s information to date other manufacturer’s products….nonetheless, it is very likely that most manufacturers followed suit at or about the same time.

Armed with this information and a quick look at one of my old transformers that has the curious footage chart on the data plate I do not see the separate grounding connection.  This is a model 721-161 (note the first “1” in the tag is scratched but upon close inspection is still legible) that has the footage chart opposite of the end with the primary power input terminals but does not have the ground lug as stated in the 1946/47 catalog.

Therefore I am placing this model as being made prior to WWII.  Since the catalog makes no mention of the footage charts being printed on the tags, it is possible that they dropped this after the war.  This assertion is going a little out on the limb, but seems likely and until I locate a similar unit to my earlier one that has BOTH the footage chart tag AND the ground lug I have no definite way to dispute this guess.  Archaeology, even neon archaeology, is often a series of educated guesses and like all such pursuits may be later disproven.

Below is a scan of the catalog for those wishing to peruse its contents for more information.

Jefferson Neon Trans Catalog 46-47

 

Lastly, if you have any similar types of catalogs from other manufacturers I would be most interested in obtaining them to further this project.  –Thanks!

 

 

****Update #3 — 5/20/2018 –****

The above transformer is a Thordarson unit, most likely of the early to mid 1930’s.  Note that the high tension bushings are attached with a flanged plate screwed into the case rather than the usual compression from the central thru-bolt.  Very high end quality, much the same construction technique used on power pole distribution transformers but in a tiny scale.  Thordarson was one of the greats, it is no wonder that like Jefferson and Sola, their old ones are often still working.

A close look at the tag for this one might lead you to believe it was made in 1928.  The serial number could easily confuse someone familiar with the usual modern date code practice.  But 1928 is not correct.  The patent numbers referenced, 1777256 and 178422, were issued in September and December of 1930, respectively.  Thus, this unit had to have been built after that time, likely from 1931 onward but due to construction technique probably before 1935 or 36.

The patents that the transformer was built under were both assigned to Jefferson Electric.  Patent 1777256 was filed July 16, 1928 by Daley, Goddard, and Sola and was granted on September 30, 1930.  Patent 1786422 was filed by Daley and Goddard on November 11, 1927 and granted on December 30, 1930.   Without dissecting a lot of minutia, we can simply state that these two patents describe what is commonly known as a laminated core neon transformer with a magnetic regulating shunt and a secondary whose midpoint is grounded.  Pretty much the same as all of them made until the modern electronic ones came about.

This brings us to a rather rare find…  A transformer that predates the issuance of those patents.

Jefferson Electric was founded in 1915, named after Thomas Jefferson-who was also an inventor.  In 1928, they merged with Chicago Fuse Manufacturing Company (founded in 1889) and formed the Chicago-Jefferson Fuse & Electric Co.  In 1931, they consolidated most of the plants into one location in Bellwood, Illinois and shortly afterward the company then became known as Jefferson Electric.

This transformer is especially rare for two reasons.  First of all, it was tagged as being made by Chicago-Jefferson and with the pre-Bellwood address, which places it firmly into the 1928-1931 timeframe.  In addition, the tag states “Patent Applied For” ….and, as its design is most likely associated with the same two patents previously mentioned, would mean that it was built sometime after the 1927 and 1928 application dates but prior to the patent issuance dates of Sept and Dec of 1930.  In addition, this model appears in the 1928 catalog.  It is therefore a safe bet that this one was made in 1928 or 29, but possibly as late as August or Sept of 1930.

Two other oddities about that tag.  There is a serial number, unusual for a Jefferson transformer.  Sometime after this was made and yet before WWII, Jefferson stopped placing serial numbers on their neon transformers.  This is the only vintage Jefferson I have encountered that had a serial number.  It does not, however, coincide with any obvious date code arrangement…so the patent applied for message and the catalog are the closest production date ranges I can apply to it for now.  The next item of interest is that this is a two part tag.  The manufacturer’s name information in one tag and the specifications below it in a separate tag.  Possibly this was done to allow for some flexibility in manufacturing:  If some common cases were used between two or more models, they could be prepped and tagged with the first tag and then have the specs tag added once it had been assigned to a specific model in the production sequence.  But more likely it could have had to do with the merger between Jefferson and Chicago Fuse.  Maybe Jefferson cut off part of the pre-made tags they had in stock and replaced them with the upper portion showing the new name after the 1928 merger and did this until the old stock of tags had been used up?  Some of the surrounding paint markings as well as the different aging of the paint of the tags themselves makes this plausible.  In either case, this was a short lived practice, likely ending by the time of the consolidation into the Bellwood plant if not before.