We assumed the striking and chiming control mechanisms were the same, however we found a subtle but important difference. At first we were unable to get the mechanism to run correctly. It just didn’t work from any obvious starting point of the two pin pinion.

The significance of the two pin pinion is that the ratchet wheel does not move in a continuous manner but rather in steps (two per revolution of the arbour). The equivalent wheel on the chiming train moves continuously.

This is important because the ratchet wheel, which contains the lifting pin, must move at the start of the striking process (to allow the horizontal arm to drop) and at the end (to lift the arm back up). For the striking train it moves one step per hour being struck. Considering the striking of one o’clock, where the ratchet wheel only moves one tooth, it is clear that there must be movement of half a tooth at the start and half a tooth at the end of striking. That means the rest position of the ratchet wheel must be mid-movement, so the spring pawl cannot sit in a fully engaged position at rest, it needs to sit half way between teeth. This doesn’t seem like a natural resting position, but it does work. When at rest one of the pins is engaged with the ratchet wheel and when the pin comes out of mesh the pawl is engaged. The result is that either a pin or the pawl is always engaged with the ratchet wheel to prevent it moving inappropriately.

The fly arbour rotates five times per tooth of the ratchet wheel / strike of the bell.

To strike one o’clock:

• The first rotation of the fly arbour moves the ratchet wheel half a tooth, the two pin pinion disengages from the ratchet wheel and the spring pawl engages. This movement is enough for the horizontal arm to drop off the lifting pin on the ratchet wheel.
• The fly arbour then rotates freely for four turns, the two pin pinion remains out of mesh with the ratchet wheel.
• In the final half rotation the second pin in the pinion engages with the ratchet wheel and moves it half a turn. This causes the lifting pin to raise the horizontal arm and stop the striking mechanism.
• The rest of the mechanism works just like the chiming mechanism.

The lifting projection has to be very slim so it isn’t able to accommodate much of a ramp – the pictured projection doesn’t look like it would work but it does, in fact it works very well. Note the pictures showing the resting positions of the ratchet wheel, pawl and two pin pinion. There is also a picture showing how the positions of the blocks were prototyped in wood (attached with double sided tape).

The control mechanism for the chiming train is now complete (all but for pinning the rotating arm to the fly arbour). Read the previous two posts on this mechanism for background.

I’ll repeat the explanation (improved and modified slightly) from the earlier post of how the process works, with pictures of the finished product.

Starting position before chiming

The horizontal arm is held just above the rest pin. It is held there by the downward projection from the arm which has been lifted onto one of the four pins around the large wheel attached to the barrel.

The rotating arm on the fly arbour is always trying to rotate clockwise but is obstructed because the square block projecting from it is locked against the square block projecting from the catch plate. Note the horizontal arm is above the rest pin.

Chiming process

The horizontal arm is lifted (note the increased gap between the rest pin and the arm) briefly by the going train, on each quarter hour. This unlocks the rotating arm, attached to the fly arbour so its projecting block can pass below the block on the catch plate.

The projecting block of the rotating arm now briefly locks against the upright (angled) block on the catch plate, awaiting its final release.

The lifting mechanism from the going train disengages, causing the horizontal arm to drop back to it’s starting position (still above the rest pin). This unlocks the rotating arm again and allows the projecting block on the arm to pass over the upright block.

The fly arbour begins to rotate, which means the rest of the striking train moves too.

About half way through the first rotation of the fly arbour the large wheel on the barrel has rotated far enough that the downward projection from the horizontal arm drops off the pin.

The horizontal arm drops down, below it’s starting point, and finally rests on the rest pin. This allows the rotating arm to rotate freely, it’s projecting block passing over the horizontal block on the catch plate.

When the chime barrel reaches the end of the tune, the next pin on the wheel reaches the point where it raises the horizontal arm.

The block on the rotating arm now collides with the horizontal block on the catch plate causing the arbour, and rest of the train, to stop abruptly (except the fly, which is able to come to  rest gradually thanks to its ratchet drive).

The completion of the flies meant the lantern pinions could be attached to the fly arbours. They are now pinned though the centre section, just like the originals, and the pins have been machined perfectly smooth with the pinion body.

The pins have been cut to length and hardened. Then the brass end cap was attached – a simple tight push fit. The end cap has then been machined to a perfect match with the body and a little bevel put on, which helps to mask any visible line at the join.

Compare the pictures of the two new pinions with one of the originals when I originally discussed their construction. Note the bronze dot just visible between the trundles where it is pinned to the arbour.

The maintaining power is now completely finished. The lever has a weight. The pawl to advance the great wheel has been made and attached (welded to the arbour). There is even a stop to prevent the lever being lifted too high and to prevent it dropping below the level of the winding peg.

From the fully charged position (lever all the way up) there is approximately four minutes to wind the going train. It actually takes about six minutes to return to resting position but because it passes in front of the winding peg it would rest on the winding handle if it were still attached beyond four minutes. Passing in front of the winding peg is deliberate, it means you have to charge it before you are able to wind the going train. You don’t want a turret clock slowing down on you else you employees might be late for work!

As we currently have no weights to wind up we have no idea how long winding will take. Luckily if it takes more than the time allowed the maintaining power can be recharged mid-wind as many times as needed.

The pictures show the pawl in three positions, from fully charged to resting gently on the teeth of the great wheel as it rotates beneath it (which it always does when not charged).

The flies now have arms and are essentially complete. The exact sizes are estimates, based on clocks seen in person and online. It may be that some tweaking of the paddles is needed further down the line to correctly control the striking. We would still like to see more turret clocks in person to make this kind of educated guesswork more accurate.

Due to the position of other components on the outside of the frame the striking fly has a crank in it. The chiming fly only needed a 1/4 inch offset to give sufficient clearance. It is normal for the two flies to be different, they aren’t both made cranked just for the sake of matching.

The arms are made from 1/4 inch steel, welded together rather than a single piece (otherwise quite a large, but narrow, plate would have been needed). The blades of the fly are made from approx 0.064 inch thick steel. Pawls have been made to engage with the previously made ratchet, with springs attached to the arms.

The arms have been sprayed in red oxide to prevent corrosion, this is not the intended final colour. We do plan to fully repaint the clock once complete but the final colour hasn’t been decided yet. Current thinking is a dark green (in which case probably British Racing Green). We don’t know what the original colour would have been, quite possibly the same blue as it came but that’s not certain.

It’s been pretty cold again here so progress had been limited for the last couple of weeks, but another small job has been completed. All three of the barrel ratchet pawls were missing from the clock, along with their springs, and the pin to attach the pawl on the going train (the pins on the other trains did more than just hold the pawl so had been left on). When the clock was directly motor driven theses were no longer needed.

Replacement parts had to be made. The springs are riveted to a small block that bolts to two threaded holes on the wheel (3BA thread), as you can see in the pictures.

When purchased the barrel and the wheel were stuck together, I was worried they might have done something nasty to join them. However,  with dismantling and cleaning, they have freed off and now work nicely as they should.

The maintaining power is a device for keeping the clock going while you wind it (and winding a turret clock can take a while). The type on this clock involves a weighted lever which must be lifted to allow the winding handle to be attached. Lifting the lever engages a pawl with the great wheel to apply force to the going train. The lever slowly returns to its resting position as the clock continues, by which time the winding should be finished.

The hole at the front of the clock for the maintaining power arbour had been covered by a John Smith & Sons, Derby plaque, which was our only clue to the fact that Smith’s had done the conversion. Smith’s confirmed this was how it used to be done when my Father visited them.

You can see in the pictures some liquid weld filling three of the bolt holes where the plaque was removed (the head sheared off the fourth). They look a bit ugly now but once they’re smoothed down and the clock is repainted you’ll never know there’d been a plaque there.

A new blush was made for the front and there isn’t one at the back (the arbour goes straight into a hole in the frame). There isn’t a great deal of movement here, but without a removable bush at at least one end there would be no way to get the arbour in and out.

Then a cranked lever is attached to the squared off arbour and a fancy brass nut holds it all in place.

Next steps are the pawl, the weight on the end of the lever and a stop to prevent the lever going down too far.

We’ve got a little bit more work done on the strike controlling mechanism (I previously called it the strike stopping mechanism, but of course it’s just as involved with starting the strike as it is with stopping it). The two “catch plates” (I don’t know if there is a proper horological term for these) had been removed. New ones have been made, but are not yet complete. To complete them they need two rectangular projections that interact with a third projection on a rotating arm attached to the fly arbour. We have also made the first of these rotating arms (minus it’s projection).

My father has seen this in action on his visit to Smith’s, and we have videos, but we still need to do some work on figuring out exactly how it works. An explanation of our current understanding follows…

Starting position before the clock begins to strike:

• The horizontal arm in the first picture (with the new plate attached) is held just above the rest pin (visible just above the right edge of the lantern pinion). It is held there by the projection you see from the back of the arm (incomplete) being lifted by a pin on the wheel behind the frame. The 12 pins on this wheel are spaced increasingly far apart to control the number of strikes.
• The rotating arm (third picture) on the fly arbour is always trying to rotate clockwise but is obstructed because the square block projecting from it is locked against the square block projecting from the catch plate.

The striking process:

1. The horizontal arm is lifted briefly by the going train, on the hour.
2. This unlocks the rotating arm attached to the fly arbour, so its projecting block can pass below the block on the catch plate.
3. The fly arbour begins to rotate, which means the rest of the train moves too.
4. Within the first rotation of the fly arbour the wheel with the pins at the back of the clock rotates and is no longer trying to lift the horizontal arm.
5. The horizontal arm drops down, below it’s starting point and rests on the rest pin.
6. The rotating arm is now able to rotate freely, its block now passing above the block on the catch plate.
7. The clock strikes the required number of times.
8. A pin on the wheel at the back lifts the horizontal arm.
9. The block on the rotating arm collides with the block on the catch plate causing the arbour, and rest of the train, to stop.
10. The fly, which is driven by a ratchet, continues to spin for several seconds, until it comes to a natural stop.

This explanation will probably be a lot easier to follow once we’ve completed it and have more pictures, or better yet a video. In the mean time these videos might help: a 1912 Smith clock in the US & the 1910 Smith clock at Trinity College, Cambridge.

The fly is a kind of air brake that controls the speed at which the connected arbour can rotate. This in turn controls the speed of chiming or striking. Obviously the timing here isn’t nearly as critical as the going train. The fly goes on the same arbour as the lantern pinions made last week. Unfortunately this whole section of the train had been removed when the clock was converted to electric drive.

The fly on a turret clock can be pretty large – we think ours needs to be up to 28 inches in diameter. When the arbour stops turning (which happens suddenly) the fly needs to able to come to a stop gently. To enable this the fly is driven through a ratchet – when the arbour stops turning the fly can continue until it stops naturally.

Our new arbour is made of medium carbon steel and will eventually be hardened at the ends where is passes through the bushes. On one end of the arbour we have the new lantern pinion. On the other end the arbour extends out through the new bush and through a loose bush to which the fly will be attached. The arbour is then squared off to take a ratchet wheel. This wheel turns with the arbour and forces the fly to turn using the spring loaded pawls attached to the fly.

The pictures show the new arbour with lantern pinion (it’s not yet attached to the arbour so the pins haven’t been trimmed and the end hasn’t been closed), the new rotating bush and ratchet wheel in position.

A bit warmer weather has meant it’s been possible to be working  in  the workshop again, so hot on the heals of the new lantern pinions yesterday comes a new bush for the fly arbour.

Rather than making it out of phosphor bronze, as you might normally for a bush, it’s machined out of brass. This is primarily because we didn’t have any phosphor bronze of sufficient size and it’s much more expensive to buy than brass. It’s a fairly substantial bush and the forces through it shouldn’t be too high (there is no force pushing the arbour against the wall of the bush, except for a little gravity), so it should be fine. If we’re wrong it’s not the end of the world, it’ll just have to be replaced when it wears.

The bush is a tight push fit into the arbour support pillar and was pressed in with a vice. Next it needs to be reamed out, using a reamer mounted on an arbour passing through the bush on the opposite pillar, to ensure perfect alignment.