[NOTICE: This Notebook is unfinished! It's reasonably complete insofar as the discussion of the actual choker valve and its bushing are concerned. It doesn't yet begin to cover the system of levers controlling the choker valve, or its cam. I've also written this Notebook hastily, to supply an immediate need. It could do with much improvement.]
This Notebook describes, in semi-technical terms intended for the layperson generally interested in the machine, the choker valve, the typemetal flow through it, and its controlling system in the Monotype-Thompson Type-Caster.
The choker valve, both as an individual component (pair of components, really) and as an entire system is probably the most intricate single subsystem of the Thompson. It is certainly the most troublesome. It can also be very confusing when learning the machine.
(But if you're trying to understand how chokers operate in other typecasting machines, then you probably want to study them and ignore the Thompson. The choker mechanism on the Monotype-Thompson is an unusual implementation of the concept of a choker, the mechanics of which are not representative of choker valves generally.)
The choker valve itself was re-engineered at least twice during the production of the Thompson. Most of the discussion here will illustrate the later system used for most Lanston Monotype Machine Company (LMMC) Monotype-Thompsons. This consists of Lanston assembly Xa7TC Choker Valve Group (which in turn is composed of a7TC1T Choker Valve and a7TC2 Choker Valve point) together with Lanston p/n a37TC2 Melting Pot choker valve bushing. The illustrations here will be of gas pot machines, but the choker valve is identical on electrically heated pots.
The term " choker valve" (or more simply " choker") is a general term for that part of a type casting machine which cuts off ("chokes") the flow of typemetal from the pump (piston) to the nozzle/nipple. Note that this function is in principle distinct from the mechanism which allows the pump well under the piston to re-fill with typemetal after a casting stroke. In the Thompson, these two functions are combined in a single mechanism, but in other machines they may be separate (machines without a choker still need a method to allow the pump well under the piston to re-fill).
The term "choker" is a functional one which describes a capability of the machine. It does not imply any particular implementation of this capability. Different type casting machines may have very differently engineered choker valves.
Not all type casting machines have choker valves. The very first pivotal type casters by David Bruce, Jr. did not. The ex-American Type Founders Company "giant pivotal" (so called; this is an informal name given because the machine is huge) owned by Gregory Jackson Walters (with which he cast the 120 point Goudy Cloister ornaments in 2014) does not. The Foucher type casters the design of which dominated European type casting from 1878 on do not. (But the Barth type caster, which was derived from the Foucher, does.) All that is necessary to avoid the use of a choker valve is to raise the nipple/nozzle above the level of the molten typemetal in the pot.
The choker valve was invented by John I. Sturgis in 1849 (US patent 6,243). Its invention postdates the first successful type casting machines (Bruce's pivotal type casters of 1838 and 1845). David Bruce, Jr. specifically acknowledged Sturgis' priority in his 1874 memoir later published as History of Typefounding in the United States (p. 61 of the 1981 Eckman edition of this work).
Note on names: One 19th century American source, Alfred Pye's article "Typefounding," Part IV in The Inland Printer, Vol. 3, No. 4 (January 1886): 203-204 refers to the choker as a "joker." This is clearly just a phonetic mistake ("joker" sounds almost like "choker"), but it is not clear whether the mistake was made by Pye (a journalist) or if he was accurately reporting contemporary workshop terminology. Legros & Grant, in Typographical Printing Surfaces (1906) , p. 274, take this one step further and refer to the choker as a "jobber." At present - until I find confirming evidence - I must regard this variant on the term as problematic. Although this work has long been regarded as a definitive source, the closer I look at Legros & Grant when they discuss machines other than those of their own manufacture, the more I find inaccuracies.
Note on sources: The discussion of the choker ("jobber") in Legros & Grant's Typographical Printing Surfaces (1906) is not useful. They seem not to grasp the purpose of the device. Curiously, the several detailed drawings of type casting machines in Wilkes' Das Schriftgießen (1990) do not show choker valves at all.
The choker valve is a system of moving components which must run inside a bath of circa 700 degree molten typemetal and must control the flow of that metal under pressure through small passages. It is no surprise, therefore, that it is typically one of the more troublesome aspects of a machine. The question of why a typecasting machine might be designed with a choker valve at all naturally arises.
The reason is that in a machine without a choker valve there is necessarily some length of passage for the typemetal between the pump mechanism and the nozzle/nipple which in a resting state is filled with air. In a choker-less machine, when a casting stroke occurs this air must go someplace. Some of it certainly escapes through spaces between machine and mold components. But since typemetal can also escape, and flows easily through spaces of well under 0.000,5 inches (0.01 mm) in size, a good machine will be as tight as possible and thus this path for escaping air is limited. The other place the air can go is into the type itself. This contributes to porosity in the cast type (bubbles of air of greater or lesser size). In all casting - of type or any other material - porosity is a serious problem.
In a type casting machine with a choker valve, this air-filled passage can be eliminated. Everything behind the choker is solid, air-less molten type metal. In front of the choker is only the mold. So the only air to worry about is that air which necessarily fills the empty mold. (There have been attempts at vacuum type casting systems, but they have not been successful.)
(One might suppose that a further benefit of a choker valve would be a reduction in the length of the passage to the nozzle, making it easier to keep it up to temperature. This is suggested, for example, by the examples of the Linotype and the Ludlow (both of which do not have choker valves, and both of which must have additional heating in place to keep this passage - called in them the throat - up to temperature). Perhaps this is so. On the other hand, in the Thompson it is still necessary to adjust the burner of gas pot machines such that the nozzle is heated by them, and it is still common to get solidification in the nozzle right at the choker valve.)
All this having been said, there is no single factor in a machine which makes it good or bad. "System Foucher" machines (by Foucher, Kustermann, or others) have no choker valve and are fine typecasters which can produce excellent type. Thompson and Barth type casters have choker valves, and are also fine typecasters which can produce excellent type. All of these machines can also be run badly to produce poor type.
So unless you're designing a new typecasting machines from scratch, the presence or absence of a choker valve doesn't really matter. The question is a matter of armchair philosophy and engineering, not shop practice. Your machine will either have one or not, and there's nothing you can do about it.
The Thompson choker valve considered in its most limited sense as a pair of components is buried inside the machine and is almost entirely invisible. The choker valve considered as an overall system is threaded throughout the entire machine and is difficult to separate from it. (A common "trick question" in Thompson instruction is "what is the other thing that the pump stop does?" The answer is that it disengages the choker too.)
In order to locate the choker valve in the machine, we must first locate the nozzle. (In other casters, particuarly pivotals, the nozzle is often called the "nipple." The Thompson is perhaps a more prudish machine.) Here's a view of a Thompson with the pot swung back (I'm in the process of emptying it) and the locations of the nozzle and the barely visible tip of the choker valve shown.
Here's a close-up view of a nozzle by itself. (Truth in advertising: These are three views of a TTMC H-575 nozzle from the spare parts for machine s/n 130. But there was no change to the Thompson nozzle during the entire production run, so this part is identical to the LMMC 62TC1.)
The Nozzle is hollow and has a tapered shank. (It is installed by driving it into the machine with a hammer and a special Nozzle Driving Tool.) The hexagon lets you twist it off when replacing or re-seating it. The rounded end fits into the Nozzle Plate which interfaces the Nozzle/Pot to the Mold. The choker fits inside the Nozzle and the tip of the choker closes off the round hole in the end of the Nozzle.
The choker valve and its bushing are hiding behind the nozzle. It is possible to see them if you empty the pot, but you have to know what you're looking for. So here is a preliminary view of the choker valve and choker valve bushing, shown in approximately the same orientations, together with the nozzle. (These pictures were taken at different times, so they don't quite line up.)
1. The Choker-valve bushing completely seals off the output of the Pump Well (on the one side) and of the bore in the Pot leading to the Nozzle (on the other side). The only path out of or in to the Pump Well is via the Choker Valve Bushing. The only path in to or out of the Nozzle is the Choker Valve bushing.
The Choker-valve Bushing is a press fit in a bore drilled/bored in the Melting Pot. Replacing it, assuming you can find or have manufactured a spare part, is nontrivial. (Cleaning it if it becomes jammed is also difficult. Doing so involves removing the Nozzle and drilling/reaming out the various holes in the Bushing, taking care not to drill too far.)
Here are some photographs of an LMMC Xa7TC Choker Valve group (consisting of one a7TC1T Choker Valve and one a7TC2 Choker Valve point, supplied assembled) and an a37TC2 Melting Pot choker valve bushing.
Below left is a view of the pump well side of the bushing. The three holes at the bottom go all the way through the bushing longitudinally. What looks like a small hole at the center is just the center-drilled blind hole for holding the part while it is being turned on a lathe. It has no function in the operation of the part.
Below right is a view from the nozzle side. You can see the three through-holes coming out on this end. You can also see the larger central hole for the choker valve itself. This central hole does not go all the way through. The slot along the top of the bushing joins up with this central hole.
Here's a view with a pica stick for approximate scale (inches on top, picas on bottom) and a set of three views showing that the small holes go all the way through but the large central one does not. (The through-holes are so thin that I couldn't get a close-up view which showed light through two of them at the same time.)
Because it's hard to show things clearly with photographs, I've constructed a 3-D model of the choker valve and its bushing. The model is freely available (this is Open Source Hardware). For more information, see the section on Groups 7, 37, and 62 - Choker, Bushing, and Nozzle in the Notebooks on Reverse-Engineering the Thompson Type-Caster.
Please note that this 3-D CAD model is at present refined only to a degree sufficient for illustration, not for the production of new components. The model of the choker valve itself is reasonably accurate, because I have several to measure. The model of the nozzle has some accurate dimensions, but because all of my nozzles are presently on machines, other parts are done by guess and by photograph. The model of the bushing is also done by photograph, as I've misplaced my notes with the measurements of it. The partial pot is just sketched in for purposes of illustration. It is generally similar to a real pot, but not correct in any detailed or dimensional aspect.
The first requirement is a consistent nomenclature for the features of these parts (where "feature" is used in the engineer/machinist's sense of "an identifiable aspect of a part," not the marketing sense of "something good"). Here's a view of the choker valve bushing identifying the features important in its operation. I should emphasize that there is no standard nomenclature here (except perhaps for "Pump Well," "Nozzle," and "Valve"). You won't find these terms in the manuals or the parts lists. I've just invented these terms because I need to call these features something.
(The illustration above does not call out the annular grooves at each end of the part or the diametrical relief in its center section. These features are important only when driving the bushing into the pot.)
I'm calling the two long grooves on either side of the valve "passages," even though they are not enclosed tubes. When the valve is fitted in its bushing, they do form passages from the pot to the area around the nozzle.
The choker-valve is a two-way valve (I haven't mentioned this yet) and so has two valve faces. When the choker is open (typemetal can come out of the nozzle) the choker valve is retracted and a conical "Valve Back Face" bears against the choker valve bushing. When the choker is closed (nozzle sealed) the choker valve is advanced and the small conical and cylindrical "Valve Face" at the end of the point bears against the inside of the nozzle. More on this later.
The next thing to do is to assemble these two components. The choker valve (the valve proper and its point, which for convenience I'll henceforth just call the "choker valve") slides in the central bore of the choker valve bushing. The Slot in the choker valve bushing is always open to the melting pot. The position of the choker valve can be controlled by engaging a lever into the Notch in the choker valve (this lever isn't shown in the illustrations here).
The choker valve has two extremes to its position. In one (shown at right below) the valve is as far into the bushing as it can go. The Valve Back Face is tight against the bushing. This cuts off the potential flow of typemetal between the main body of the pot and the area in front of the bushing. This position will be called the "open" position. This may seem counterintuitive now (as in it the choker is closing off the central bore of the bushing), but in the larger picture it makes sense. It is the position in which the choker valve is retracted and the nozzle is open, allowing metal to exit the pot.
The other extreme position is shown below left. In it, the choker valve is extended out as far as it will go (its motion is limited by hitting the nozzle, which isn't yet shown). In this, the "closed" position, typemetal can flow freely from the main body of the melting pot through the slot on the bushing and into the area in front of the bushing.
That explains the passage of typemetal through the Slot and the Valve Bore (the large central bore) of the bushing, but what about the three "Through-Holes"? The answer is simple: they remain open at all times, regardless of the position of the choker valve. Typemetal can always pass freely through them from one side of the bushing to the other.
It's time to put the bushing and valve in context. Below is a cut-away view of the pot (highly simplified!) with the bushing, choker, and nozzle in place. The term "Pump Well" is attested on p. 55 of the Lanston 1950 Monotype-Thompson Adjustments manual. The cylindrical part of the pot (cast integrally with it) which has the choker valve bushing on one side and the nozzle on the other side has no standard name that I'm aware of. It's important to this discussion, so it needs a name. I'll call it the "nozzle boss." (In traditional engineering, a "boss" is a raised part of a casting (think of "embossed" lettering, which is raised lettering). From the outside of the pot, at least, this feature is a boss.)
With the piston in place (assuming it is well-fitting) there is no way other than these three bushing through-holes for typemetal to get into and out of the pump well. All typemetal during the casting stroke leave through them, and the pump well is refilled with typemetal after the casting stroke through them.
The other end of the bushing is harder to view because it's all hidden within the part of the pot casting I'm calling the "Nozzle Boss." You can render the pot as transparent, but the rendering the nozzle is harder due to its complexity. Here, to start, is a view from an angle with transparent pot and nozzle.
The first thing to notice is that the choker valve bushing and the nozzle cap each end of the bore through the nozzle boss. This forms an enclosed space that needs a name (there is no standard name for it). I'll call it the "Nozzle Chamber." Typemetal can get in to or out of the nozzle chamber in any of three ways:
(Comment: The end of the choker valve appears the same color as the nozzle in the image above because it is not a true sectional view but rather a side view with the nozzle rendered translucently. The end of the choker valve is picking up the coloring of the translucent nozzle. But I've just colored in the nozzle chamber in solid red by hand.)
So now we'e seen all of the parts of the choker valve (with its bushing, its seating in the pump well, its seating in the "nozzle boss", and the nozzle. We've seen the function of each part separately. What would be very useful now is an animation of it all in operation. Unfortunately, my CAD skills and tools aren't up to that yet. So we'll have to use a thousand words instead of a single picture. The important things are the paths of the typemetal and the sequence of operations.
There is a continuous path for typemetal from the main reservoir of the Pot, through the Slot in the top of the Choker-valve Bushing, through the two Passages in the Choker Valve, to the Nozzle Chamber, though the three Through-Holes in the Choker-valve Bushing, into the Pump Well.
When the choker valve is open, the path from the Nozzle Chamber to the main reservoir of the Pot is closed. Once the Piston begins its stroke, the typemetal in the Pump Well / Through-Holes / Nozzle Chamber is under pressure.
7. The piston is spring-loaded. Once the mold has filled with typemetal, it should continue to push down on the typemetal in the pump well. This keeps the typemetal pressurized until the type and jet solidify in the mold.
Several of the patents issued to John S. Thompson describe machine features that either anticipate or are part of the Thompson as manufactured. At least three of them (disregarding overseas filings) describe choker valves. However, none of these correspond to machines as produced (at least from s/n 130 onward, the earliest machine for which I have information) or their documentation. Still, they're interesting.
US Patent 1,029,243 (1907/1912)
US patent 1,029,243, "Metal-Pot for Type-Casting Machines and the Like" [and choker valve]. Filed 1907-09-20. Issued 1912-06-11 to John S. Thompson. Not assigned.
Local copy: us-1029243.pdf
By filing date, John S. Thompson's September, 1906 patent 1,029,243, "Metal-Pot for Type-Casting Machines and the Like," is not his earliest patent describing a choker valve (that would be 1,119,733, filed 1906-01-06). However, it is probably the most robustly simple of all those considered in this Notebook. Note especially the very direct path for typemetal from the Pot to refill the "Pump Cylinder" (as he called it in the patent).
US Patent 1,119,733 (1906/1914)
US patent 1,119,733, "Type-Casting Machine." Filed 1906-01-06 as application serial number 294,958. Issued 1914-10-02 to John S. Thompson. Assigned to Thompson Type Machine Company.
This is a PDF created from lossless PNG conversions of TIFFs downloaded directly from the US Patent and Trademark Office website.
Local copy: us-1119733.pdf
This patent, which is by filing date the earliest I've yet found by John S. Thompson which describes a choker, shows a mechanism which is closer to that of early production Thompsons, but not quite the same. Note in particular the way in which the back of the choker valve closes off the "Reservoir" (Pump Well) by sliding into a bore rather than by seating on a conical seat as in later practice. This machine also has a curious long nipple with a flat face.
This patent is also interesting because in it Thompson says, without explaining why, that "preferabley ... when in position closing the nipple, the tip of the choker will extend slightly beyond the face of the nipple." This is an early confirmation by Thompson himself of a hypothesis entertained by Sky Shipley, of Skyline Type Foundry. The Thompson was supplied with a choker valve the point of which was over-long. It was intended that the operator would grind this to length. Typical practice has been to grind it flush with the end of the nozzle. However, in operation with the nozzle plate in place this leaves a void, the thickness of the nozzle plate at the nozzle, between the nozzle and the back of the mold jet block. It is common to get a bit of frozen typemetal in this space. Shipley's theory is that the choker valve should be ground down so that it projects beyonod the nozzle by this thickness. Apparently John Thompson agreed.
US Patent 1,026,185 (1907/1912)
US patent 1,026,185, "Type Casting Machine." Filed 1907-12-05 as application serial number 405,230. Issued 1912-05-14 to John S. Thompson. Not assigned. The Thompson as manufactured begins to emerge in this patent.
Local copy: us-1026185.pdf
This style of choker valve is attested both by spare parts for TTMC Thompson Type Caster s/n 130 (a machine I own which is currently in a friend's shop) and by illustrations in the 1916 Instructions to Operators of the Thompson Typecaster.
It differs from the valve shown in Thompson's patent 1,026,185 (filed 1907) in the interface between the back end of the choker valve and the pump well. In the 1907 patent, the back end of the choker slides into a bore leading to the pump well. In the 1916 manual, the back end of the choker seats on a conical seat which is in turn the face of a bore leading to the pump well.
(The nozzle as drawn in the manual differs from the one in the spares for TTMC machine s/n 130. The one in the manual has a cone on its interior, while the one for s/n 130 has a hemispherical surface on its interior. This may be an error in the manual. In any event, it doesn't matter since the choker valve seals the nozzle by its point in the hole at the end of the nozzle. The shape of the rest of the interior of the nozzle is not important.)
The 1925 Thompson Type Machine Company manual shows a different mechanism for the choker valve itself - a complete re-engineering of the device. It's now a two-part mechanism consisting of p/n C-531 Assembled Choker. (Parts not sold separately) and C-580 Choker Valve Bushing. If later practice is a guide, the "Assembled Choker" is just the Choker Valve plus its removable tip.
At some point after manufacturing either 248 or 338 machines, the LMMC changed the design of the choker valve. The change wasn't dramatic (I can't tell the difference between the old part as shown in the 1925 illustrations and later parts) but it was enough to cause Lanston to issue new part numbers. Both the bushing and nozzle were prefixed with a revision letter: a7TC1T, a7TC2, a37TC2. The nozzle remained unchanged.
There is some ambiguity in the 1942 Parts Price List as to when the LMMC made the change to the new choker valve and bushing design. The earlier (late TTMC / early LMMC) version of the valve itself is LMMC p/n 7TC1T Choker Valve, furnished assembled with 7TC2 Choker valve point. This was supplied from machine s/n 10051 (which we presume was the serial number assigned to the first LMMC Monotype-Thompson) through s/n 10299. So far so good.
Note also that the new style assembly (valve plus bushing) was "plug compatible" on the assembly level as an old. Once the new style was introduced, the LMMC declared the old style obsolete and supplied the new style in its place. So any particular surviving TTMC or early s/n LMMC machine might well have the new style bushing/valve installed.
All of the photographs here which were taken by me, and their annotated versions, are licensed under the same Creative Commons license as the page in general.
The CAD models of the Thompson developed by me are Open Source Hardware. See the Notebook on Open Source Hardware on CircuitousRoot) for more information. The 2-D images derived from these models are licensed under the same Creative Commons license as the page in general.
The 1916 and 1925 Thompson manuals are in the public domain.
The text and images of US patents are in the public domain by law.
All portions of this document not noted otherwise are Copyright © 2015 by David M. MacMillan and Rollande Krandall.
Circuitous Root is a Registered Trademark of David M. MacMillan and Rollande Krandall.
This work is licensed under the Creative Commons "Attribution - ShareAlike" license. See http://creativecommons.org/licenses/by-sa/3.0/ for its terms.
Presented originally by Circuitous Root®