Last but not least of the control valves to be serviced on my Monotype caster is what I refer to as the Mould Signalling control valve. Depending on what options the caster is fitted with, this valve might be used for casting low quads (rather than having the bridge detect the special quad mats which have a shallow cone hole for the centering pin), or for controlling one part of the mould blade when using a duplex or triplex mould. Some features require additional pneumatic mould control, so if you have a triplex mould or want low quads from a duplex mould, your caster would have a second (essentially identical except for choices of air lines) mould control valve beside this one, and for the triplex mould and the high-quad shutter used in mathematical setting, the Unit Adding control valve is also co-opted.

This valve (and its twin if secondary mould control is installed) are quite simple: Signalling the N air line not only raises the N air pin, but also diverts the 0075 signal from the 0075 air pin to the mould signalling line (secondary mould control uses 0005). A rod screwed into the top of the valve prevents it from moving if this operation is not wanted (in which case the N+0005 and N+0075 would act just like 0005 and 0075, causing the justification wedges to be set, turning off the pump, and/or causing the galley to operate, as appropriate).

At least, that is according to the ‘Monotype’ Composition Caster Manual. On my caster, the H and J lines are used instead of N and 0075, respectively, and I still don’t understand why. As a result of this, signalling the mould also raises the H air pin. If the signalling is used to select low quads, it doesn’t really matter which air pin pops up in the rear pin block, and it could be argued that H centers the matcase, producing on average a minuscule reduction in wear and noise when the matcase is positioned for the next character. But for use with a duplex or triplex mould, this limits casting to columns NI through H of the matcase when the mould is signaled (either regular or secondary).

In any case, I have reassembled this valve, putting the bent end of the spring inside the piston so it can’t jam when the piston moves, and reinstalled it on the caster. With all the control valves back in place I have to solve my other pneumatic woes: air leaks and stuck air pins.

Mould Signalling Control Valve re-installed with the top screw all the way down, disabling it.

After a short side trip to test and trace the individual air lines on my caster I have returned to the task of refurbishing the various pneumatic control valves.

The next one to do is the Quadding and Centering control valve. It actually consists of two valves stacked on each other. The outer one (on the right in the photo) is actuated by either A or B air from the paper tower, and its purpose is to prevent air from the C line of the paper tower from reaching the C air pin if either A or B is also selected. The inner one is actuated by the C air from the paper tower, and its purpose is to divert the air coming from the A and B paper tower lines to the Quadding and Centering actuators (rather than the A and B air pins) via the manifold that forms part of the mounting plate for the Unit Shift control valve. From this description it is pretty clear that none of the A, B, or C air pins can be raised for a character being cast for quadding or centering. This is not considered a shortcoming since this feature is intended for casting spaces, in fact, em quads specifically, and such spaces are typically located in the right-hand columns of the matcase, often in column O (which requires no air pins to select).

Rather than risk damaging more of the under-table copper pipes, I opted to clean the mounting plate for this valve in place, so it is not quite as sparkling clean as the plates for the Unit Shift and Unit Adding, but at least it is steel-coloured rather than gunk-coloured. I reassembled the valve using a light oil on the pistons and a heavier oil on the mating surfaces. I left the bottom plate screws just snug until I had snugged the mounting screws to hold the two valve bodies tight together, then I tightened all 8 screws.

The only control valve left to service is the mould signalling control valve.

The Quadding and Centering control valve, re-assembled and re-installed.

On removing the dummy plate on my Unit Adding valve base and testing individual air lines, I have determined that:

1. This valve uses lines I, K, N, 0005, and 0075 (not J, K, N, 0005, and 0075 as my manual states).
2. I had swapped two of the lines coming to this plate, causing my I and 0005 lines to be interchanged.

To fix problem number 2 I had to remove the mounting plate, disconnecting all the air lines, and re-mount it with the lines coming to the correct ports. I had tried just removing the two swapped lines but because of the way they were crossed I was unable to reposition them with all the other lines still in place.

Fortunately I found installing the plate much easier this time, perhaps because I inadvertently used a slightly different procedure.

On tracing all the lines I also found that my previous statement about the mould signalling (that it uses G and H) was incorrect; it actually uses H and J, which still does not match what the manual says.

I now have the rear pin block cover off and all the pins out for cleaning; once this is all reassembled and the remaining control valves are reinstalled I should find that all these airpins operate properly.

Using a sports inflator needle to apply compressed air to each air line individually, I am finding several oddities.

The first is certainly not unexpected: Several of the airpins are sticking. I think my best bet for these is to remove the two pinblock cover plates, blow gunk out of the lines, and clean and lubricate everything. Although the airpins have a hex socket in their top so they can be twisted to unstick them, every third airpin is obstructed by the rods that guide the pin jaws. I might try a ball-tip Allen key to turn these, but it is a pretty tight fit.

The next is that the I and 0005 pins appear to be swapped: applying air to one of these holes in the paper tower actuates the other airpin.

Finally, the mould signalling valve seems to use H and G together to signal the mould (and raising the H pin), but the documentation I have (the 2-binder ‘Monotype’ Composition Caster Manual) says it uses N and 0075.

I’m considering the possibility that I got some lines mixed up when I removed and replaced the bases for the Unit Adding or Unit Shift, but neither of these mechanisms uses G, H, or I (Unit Shift uses D, E, and F, and Unit Adding uses 0.005, 0.0075, J, K, and N according to the manual).

I will remove the Unit Adding dummy valve to identify exactly which lines lead where. Hopefully I can solve these mysteries.

It seems to me that the design choices of which lines control certain options never accounted for combinations of options that could have been useful in the future. For instance, the quadding and centering cannot be used to cast repeated characters from the A, B, or C columns of the matcase (it was only intended for quads, which are generally in the O column). The mould signalling uses N (at least, as my documentation states) and so cannot be used with 15×17 matcases because columns I and L would be inaccessible when signalling the mould (the N+I or N+L combinations would instead select column NI or NL). This is fine when it is used to select low quads but not when used to signal the side blade for a duplex or triplex mould; probably mats for duplex or triplex mould usage were only produced in 15×15 arrangements. Only the Unit Shift is designed in a manner that has no adverse impact on other functionality, but there is no reason why other options could not have been designed the same way.

Just for a lark, I tried connecting my Monotype caster to a compressed air supply, and found that the air connection to the paper tower leaked so badly that the (admittedly small) compressor could not keep up. I tried putting a little grease on the connection but that did not help much.

The air supply rises from below the caster table to the top of the paper tower in the tube to the left of the photo and enters into the connector, which is installed over the pivot shaft for the paper clamp. It is held on by a nut and washer which clamp it against a spacer collar. The shaft is hollow with a radial hole (facing away from the camera) to allow the air to pass from the connector through the shaft to the paper clamp. The nut can’t be tight against the connector because the shaft must pivot a bit as the paper clamp moves up and down for each cycle of the caster. I have to wonder what prevents the nut from working itself loose from this movement. I think the nut actually tightens against the end of the shaft (not the connector) and so the nut ans washer are locked to the shaft and rotate with it.

I suspect that the leakage is due to a subtle bend in the riser pipe skewing the connector a bit. Flexing the pipe causes the leakage to change dramatically.

I took the joint apart and found that the spacing collar does not appear (to me, anyway) to be long enough, causing the connector to tighten against the shoulder of the shaft instead. The lengthwise position of the shaft is determined by a groove that the pinch screw of the paper clamp (upper right in the photos) passes through so I can’t just move the shaft in a tiny bit.

I will be doing some careful measuring to try to determine what surfaces should be providing the air seal here. It may, for instance, be that the inside of the connector should be a closer fit to the shaft to provide a seal. It might be possible to find a pair of O-rings to fit the recess in either end of the connector, or perhaps a pair of thin flat resilient washers will do the trick.

This recess on either side of the connector might be able to hold an O-ring.

In any case, despite the leakage, I managed to apply some air to the system. I found that many of the airpins were stuck, and even when unstuck some still failed to rise. Several years ago, when I first started work on the caster, I had used an inflating needle to inject air directly into each hole in the table where the paper tower mounts, and the pins worked well then. This makes me wonder if the pipes in the paper tower are plugged with dust. The interior of the tower certainly had plenty of oily dust in it!

After breaking two pipe ends while reinstalling the mounting plate for the Unit Adding valve, I decided that soldering in new ends would be easier than trying to get a flaring tool into the tight spots under the Monotype caster’s table.

I used my lathe again to thread some spare pipe ends to fit the mounting hardware, and bent and cut them to match the broken pieces I had cut out from the caster. I tinned the ends that were to be soldered to make the solder job a bit easier. I also cut two short pieces of ¼″ soft copper tubing to make connectors. These were a looser fit that I would have liked, but still close enough for a solder joint to work. In retrospect I probably could have used by lathe to swage the tubing to the correct inside diameter.

I cleaned and fluxed the pipes to be soldered, installed the new pipe ends into the valve mounting plate, and slipped my makeshift couplings into position. I soldered the couplings using a small butane torch and finished reinstalling the valve mounting plate and the dummy valve.

You can see the finished solder joints and below and to the right of each the flare couplings that I used to splice in the original pipe repairs I had done.

The Unit Adding dummy valve reinstalled on its mounting plate

There are still two valves left to do: the Quadding & Centering valve and the pneumatic mould signalling valve. I hope I can get them done with no further broken pipe ends!

I put my replacement pneumatic lines on my Monotype caster and replaced the Unit Adding valve mounting plate. You can see them in the photo as the two clean pipes near the bottom, and the two shiny threaded ends second row from the top. My measurements worked out well, with the pipe ends coming out at the right locations with no force required.

To assist in fitting each pipe end through the appropriate hole in the plate I put short pieces of spring wire in each pipe and passed these wires through the matching holes in the plate. I attached the plate to the edge of the table using its mounting screws, then pushed each pipe individually to the exact spot so it would pop into its hole (the spring wire only held them in approximate position).

Unfortunately in doing this the threaded ends broke off two other pipes, so I am back to where I was with two pipes to repair and fit to this mounting plate.

Because I can’t reliably form flared pipe ends (and have no clearance here for the flaring tool anyway) I will have to use soldered connections to put replacement threaded ends on the broken pipes. I am reasonably sure that a short length of ¼″ copper tubing will work as a solder splice for this pipe.

I will have to be more careful reinstalling the other two mounting plates (for the Quadding & Centering valve and the pneumatic mould signalling). Although better that the soft iron wire I used when installing the Unit Shift valve mounting plate, the spring wire still doesn’t really do the positioning trick well. A better solution would be short tubes whose outside diameter matched the smaller diameter of the tee nuts and which were threaded to screw onto the pipe ends. Such tubes would position each tube exactly and would eliminate the need to push or pry the tube ends into position.