Monday

A Gillett & Johnston Waiting Train! Text Updated 1.6.2015

*
The Ottawa Peace Tower, Gillett & Johnston, WT. [G&J WT]

This all started when I found an image of a Gillet and Johnston waiting train on ClockDoc:


There follows one of my usual, rambling discourses on this rare and fascinating movement. As I have obtained better images and better understanding of the device, I have continuously modified both the text and images. Patience will eventually be rewarded with a more precise description of the Gillett and Johnston Waiting Train. Reload the page as often as you see fit to enjoy(?) the latest edition. 

 

Having publicised my interest in the G&J Waiting Train within the electric clock milieu several contacts kindly sent me further images. An online picture search produced only two more images so far.

Finally(?) for a good dose of alternative reality I suggest you scroll down to the Google Maps link below and take the virtual tour of the Ottowa Peace Clock  Tower.

Unfortunately, all of the images available (including Google's) show reflections from the protective, plastic, display case. This is very unfortunate because it detracts from the sharpness and contrast of all of the pictures of the available G&J "clock" movement. Though being a WT it is not strictly a clock in its own right. More a slow speed, slave motor with external synchronisation of its drive speed controlling the accuracy of its timekeeping.

This image is credited to Barry Hushner Picasaweb Albums.

Thanks to my helpful contacts I now know that this exceedingly rare and interesting movement was installed in the Ottawa Peace Tower of the Canadian Parliament building in 1927. This particular G&J turret clock may well be truly unique. The cost and prestige of such an installation would surely warrant a special effort.

I know of only one other G&J WT. It is also situated  in Canada but still working The present caretaker did not respond to my email requesting details or pictures after I found a local newspaper article online. I have now contacted the editor of the local newspaper, which carried several articles on the clock and its tower, in the hope of obtaining an image to share here. Pictures of the crown wheel cluster of the lead-off work and the G&J master clock are shown but not of the clock movement itself. Though the carefully worded description strongly suggests it is another Gillett & Johnston Waiting Train movement.  [I have since been informed that the clock keeper of the other G&J WT is both elderly and infirm.]

Rather than having a local master clock, the Peace Tower G&J WT was synchronised, at half minute intervals, by a signal from the Dominion Astrophysical Observatory. The movement was finally replaced in 1970 after 43 years of public service. A synchronous motor drive then took over timekeeping duties.

It is interesting that the Observatory also ceased to function in 1970. The lack of a synchronising signal might well have forced the retirement of the G&J WT. Though this is just mere speculation on my part it is still a possibility. Though I have seen descriptions of the "original clock" as "not working."

Image captured from Google Street View.

On retirement the G&J movement was put on display in the Peace Tower's Observation Gallery. Being housed in a clear, plastic, protective case it makes photography of the movement rather difficult. Particularly if flash is used. With windows on all four sides of the large room and light coloured, tiled surfaces all around the task of high resolution, technical photography or videography is quite a difficult one.

Ideally, it would require a dark background to be placed behind the camera to avoid all reflections from the facing surface of the plastic display case. The light from the other sides will not produce reflections and will probably light the movement quite evenly. A local photographer just needs an assistant to hold up a black sheet temporarily behind the camera to obtain far better images than hitherto. Choosing a quiet time with few visitors and a bright, overcast day for the photography would be eminently sensible.

I have used the same method to photograph clocks with glazed and even domed glass covered dials. The dark background makes the intervening glass (or plastic) simply disappear from view! Except when flash is used, of course. The photographer should also wear dark clothing to avoid their own image being reflected. While the plastic case may not be optically perfect it should not hinder really sharp images. In a perfect world the present display case would be replaced with a toughened glass box with a modern anti-reflection treatment. Much as is used for valuable paintings.

G&J but must have really struggled to design a WT without infringing on Gents' established patents. The G&J design must inevitably follow at least some of Gents' Pulsynetic WT basic functions. Though with considerable changes in the detail and geometry of the layout compared with its already, well proven rival.

In some ways the G&J harks right back to the Gents' WTs of two decades earlier. It could be argued that G&J were only chosen to install their own WT because they also provided the clock dials, carillon, no less than 53 bells and a motor driven chiming and striking machine in the same tower. It would have been extremely churlish for the architect to insist on a competitor's WT turret clock movement just to show the time!

Gillet & Johnston were world renowned turret clock makers and bell casters and even manufactured their own electric master clock and slave dials. They must have clearly seen the advantages enjoyed by the Gent's WT and its negative effect on the potential sales of all weight driven turret locks. The Gents' WT had made all weight driven turret clocks largely redundant.

Without needing a human winder and lots of very precise gearwheels, levers and bearings, an electrical turret clock could be made very much simpler,  considerably smaller for the same power and cheaper to purchase. Not to mention being very much cheaper to run and maintain and all in a very compact space.

The message was already writ large on the remaining clock factory walls following the success of the Gents' Pulsynetic Waiting Train turret clock. The latter had been employed for some very prestigious and many world record sized clock dials. Countless public dials had been placed where a weight driven clock was impossible to fit or far too difficult of access to be wound regularly. WTs were even placed on tall, slim war memorials and industrial chimneys. Without any need for weight shafts and regular access a WT could be fitted almost anywhere and (almost) forgotten. Provided with a signal from a master clock a WT automatically enjoyed the accuracy of timekeeping of that same master clock.  

From examination of the various images, shown here, I would suggest that the G&J movement's cast back plate is (very approximately) 2' (60cm) high and a little wider over the tapered sides. It is a very substantial, tray-like, sand casting with a deeply ribbed back. (See the last Google image below)

The next obvious feature of the G&J WT is the long, vertical drive shaft. This is fitted with a large, fine toothed wormwheel. The black painted castings at the very top of the shaft form a typical universal joint for the lead-off work to the clock hands. Buildings settle and move with humidity, wind and temperature changes. The lead-off work allows for these changes without causing loss of accurate drive to the clock hands.



This large wormwheel is driven by a worm on the same shaft as the count wheel. Lead-off work usually turns at one revolution per hour. The hour hand(s) are driven from the 1rph minute shaft by a 12:1 gear reduction called the motionwork. This simple gear arrangement is usually placed immediately behind the dial(s) for simplicity. This also allows work to take place on the individual dials or setting them to time if necessary. 

Quite unbelievably: Google Maps Street View of the Peace Tower Observation Gallery allows me to confirm that the G&J wormwheel has 240 teeth. (Counting 40T between two of the six spokes) Prior to this the available images had me believing there were only 180 teeth. Which made no real sense for timekeeping. 

A two start worm would reduce the effective number of teeth by two to one. [Making the equivalent of a 120T wormwheel with single start worm for rotational speed purposes.] Now we can safely assume that the count wheel rotates slightly faster than once a minute. No other speed of rotation makes sense if the lead-off work is to rotate once per hour to drive the minute hands of the clock dials.  

The ratchet-toothed, count wheel has approximately 28 or 29 teeth. It is driven, tooth by tooth, on top of the wheel, by a double pivoted gathering pawl. This pawl is pivoted to the back plate just to the left of the count wheel. The leftward extension of the drive pawl appears to be lifted on both swings of the pendulum by a wheel pivoted in the pendulum rod. A curved ramp on the underside of the pawl lifts the pawl on every swing to both the left and right.

The vertical part of the pawl rocks, while the other end of the pawl is in contact with the count wheel teeth. So the front edge of the pawl will push another tooth to the right [clockwise] on every swing of the pendulum. The Gents' WT only gathers a tooth only on every swing to the left. So the Gent's WT must use a 15 tooth ratchet wheel to achieve half minute rotation using its slightly shortened pendulum. This is to ensure that the WT mechanism is given a couple of seconds to correct the timekeeping. If the pendulum was the correct length for normal timekeeping there would be no time in which to correct it's inaccuracies.

If the G&J count wheel has 28 or 29 teeth and needs to rotate once in (say) 28 seconds then we can obtain the length of the pendulum from the number of beats required. The slightly shorter period (than 30 seconds) is necessary to allow a remote time signal to restart the drive to the clock hands by some means. (Yet to be ascertained in the case of the G&J WT) If the main plate is roughly 60cm high then the effective pendulum length must be somewhere around 70cm. Much the same as the Gents' C40A WT.

If the WT movement's pendulum kept exactly to time, by itself, it would offer no opportunity to set it right with an ultra-precise time signal from a remote master clock. Much like the Gents' WT, the G&J pendulum shows no obvious sign of being temperature compensated. So the accurate time signal remains vital to maintaining accurate timekeeping.

In case there was any doubt, the pendulum swings continuously regardless of the WT function. Without an accurate time reset, at fairly frequent intervals, the heavy pendulum, by itself, could not keep good time. It merely acts as a self-maintaining, low speed, reciprocating motor to drive the countwheel. The drive pawl turns powerful linear movement into rotational movement at the countwheel. The horizontal count wheel shaft drives the large wormwheel (via the worm) and thence the lead-off work (rods and couplings) to the clock dials. The worm and wormwheel greatly amplify the power provided by the heavy, swinging the pendulum as well as greatly reducing its speed. [By 120:1 from 2rpm to 1 rev per hour]

Though no worm and wormwheel set are 100% efficient the increase in torque must easily exceed 100 times in a 2/240 tooth (120:1) arrangement. It was this massive increase in power advantage of the WT which made weight driven clocks redundant overnight. The latter used massive weights to drive the slow end of the train, with all the friction involved in a long train of gear wheels, pinions and heavily loaded, plain bearings. Leaving very little power left at the faster end of the gear wheel train for actually driving the clock hands.

In a purely mechanical turret clock reality the clock weights had to be made considerably heavier than theoretically necessary. Just to ensure, long term reliability in bad weather conditions and in case of poor lubrication practices and lack of maintenance. As the pinions wear in the typically filthy conditions of a clock chamber they lose their theoretical shape. Which is why many tower clocks used lantern pinions. Which allowed dust to fall through. Until, that is, the pinion leaves were oiled by a well meaning but ignorant person. It is not uncommon to see additional, non-original weights having been added to weight driven clocks. The extra weights merely accelerated wear on the entire clock mechanism.

In Gents' WT the drive to the clock hands is physically paused by raising the drive pawl out of the teeth of the count wheel at half minute intervals. Though usually only for a second or two. Before being instantly lowered again by the time signal energising the relay electromagnet which released a supporting latch. It is the act of restarting the drive to the clock hands precisely on time which ensures great accuracy in the WT's timekeeping. The starting time of the pause to the clock drive is irrelevant. Provided, of course, that the drive pawl has been safely lifted and latched before the time (restart)  signal arrives. In fact the WT's count wheel can be turned forwards to lift the drive pawl prematurely without having any effect on the functioning of the WT mechanism. Though the clock hands will show the clock is slightly fast.

The worm has the further advantage of continuously locking the wormwheel against any unwanted rotation. So the clock hands aren't suddenly free to turn due to fierce winds or any imbalances from icing or flocks of birds sitting on the hands. The short pause in drive to the hands goes completely unnoticed by those glancing up at the clock dial to tell the time. A keen horologist with a pair of binoculars might just catch the second or two pause in the minute hand's small forward steps around the dial.

The G&J WT mechanism is a seemingly delicate device compared with the relatively massive gathering pawl and robust components used throughout by Gents highly successful 'Pulsynetic' WTs. However, the seeming delicacy should not be lightly dismissed as a sign of weakness. The very much larger diameter count wheel and the drive on every swing is a clever way of reducing loads on the system. Gents used a relatively small count wheel which must have placed much higher loads on the gathering of its teeth. It was thus vitally necessary to use strong and sturdy components to ensure a long a reliable lifetime of gathering teeth on the count wheel.

The G&J WT's long working life (ca.1927-70) suggests the G&J WT must have avoided any serious design failures. Otherwise the movement would have been quickly replaced. Canada, being so far north,  is renowned for being rather cold in winter. Icing and snow on the clock hands must have occurred repeatedly over the years.  An unreliable clock in such an important position as the Canadian Parliament Buildings would not have been tolerated for so long. So we can only assume it was a success. The other G&J WT in Canada is reported to suffer from icing of the clock hands. Requiring regular attendance by the clock repairer to free them and restore the local time service.

At the very bottom of the green painted movement plate, just above the heavy bob, are the Hipp Toggle and V-block. The Toggle is pivoted freely in its supporting bracket on the pendulum rod. The V-block sits on top of the Hipp contact system.  The Toggle and V-block control the rather complicated looking main electrical contacts. These contacts would switch on the pendulum drive power (briefly) only when a minimum pendulum arc was reached. When the contacts closed the pendulum would be given another push to keep it swinging for a while longer. 

The large, twin electromagnets at the very top of the movement plate provide that push to drive the pendulum. Their armature is concentrically pivoted with the pendulum support bearing. A long, downward projecting, extension arm gives the pendulum a direct push. Albeit via a stiff spring blade, with electrical contact at its tip, before the pad on the armature impulse arm itself reaches the pendulum rod. This method of propulsion is more direct and quite unlike Gent's more mechanical impulse ramp, rocking armature and impulse roller.

What purpose has the electrical contact? The pendulum drive impulses are infrequent and highly irregular. Depending entirely on the loads presented by the clock hands and associated lead-off rods, bearings and motion work. Perhaps the contact is a safety device to avoid damage to the coils if the pendulum should accidentally come to rest? I am guessing here slightly since Hipp Toggles and V-blocks are usually arranged to avoid the contacts ever remaining closed even if the pendulum drops below its usual arc. Or actually stops swinging altogether under heavy icing loads on the hands.

Perhaps the armature contacts are part of an extended drive impulse arrangement? This might explain the need for complicated levers associated with the pendulum drive switch contacts. Extended drive impulses were already familiar in some quite early, master clock designs.There appears to be a long lever which might have closed the contacts manually. Allowing the pendulum to brought safely up to its normal arc until the Hipp Toggle took over normal arc maintenance duties.Though none of the images is really clear enough to confirm the true purpose of the long arm extending to the left of the contact assembly.This may simply be a manual lever to close the pendulum contact to get the pendulum swinging again after a stop.

To the right of the ratchet wheel I can see twin, parallel strips of steel. There also seems to be a forward projection. From examination of the images available these seem to have no obvious function with regards to the nearby levers. However, the wormwheel has four triangular blocks projecting above the rim. Could these contact the metal strips? And what would occur if they did? Was it a chiming start contact system to occur at the quarter hours? We know that the quarter chimes were a feature of the clock and bell system. Gents used a separate unit to instigate striking and chiming in their own systems. As did Synchronome. It would be rather novel if G&J managed this from the tower clock movement itself.

Just below the large, double drive coils is the much smaller relay electromagnet.

The relay electromagnet's armature has a sagging tension spring affixed  The relay armature has fallen from its usual position up against the electromagnet core.

Here is a view from yet another angle of the ratchet count wheel and WT relay electromagnet. Unfortunately the camera flash has reflected in the protective plastic display case in precisely the wrong area.

The ratchet-toothed countwheel has a brass pin to lift the L-shaped arm and thence the drive/gathering pawl and backstop lever. The L-shaped arm has a limit stop mounted on the small metal block attached to the back plate. Unfortunately the armature has been fixed below the stop bracket for display. It would always have been above the stop when working normally.

Normally an armature should rest very close to its electromagnet core. The relay electromagnet cannot possibly attract an armature from the present distance and the stop bracket would only get in the way.

It can be seen that the active tip of the drive pawl forms a broad plate. (measured from front to back) Much wider than is necessary just to push the teeth of the count wheel around. This plate allows the L-shaped lever to lift the drive pawl clear of the wheel teeth by rising under the broad tip of the pawl when lifted by the brass pin on the countwheel. 

The G&J WT mechanism works thus: A pin projecting from the back of the count wheel lifts the L-shaped lever and the nearby driving pawl clear of the ratchet teeth. The L-shaped lever is latched by the face of its small block by the armature as it supports the driving pawl clear of the ratchet teeth for a second or two. Then the relay electromagnet receives its electrical signal from the remote master clock. The armature rises, unlocking the L-shaped arm from the face of the block. The L-shaped arm can now drop to lock on the armature in the notch above the block. Allowing the drive pawl back into contact with the teeth of the countwheel.  The pawl then continues driving the clock hands as normal.

The L-shaped arm obviously prefers to be in its downward position due to gravity. It is balanced to drop automatically when not being held up by the countwheel pin. 




I have now discovered that Google Street View has "mapped" the Peace Tower observation gallery with fine images!

The viewer can wander a rather fixed path around the gallery in crisp resolution and examine the WT from three different angles. Though one cannot go very close. Attempting to zoom in on the clock takes the viewer right through to the far wall! Unfortunately the display case is still highly reflective and detracting of any real detail.

Though interestingly, the Google camera trolley casts a dark reflection of itself in the plastic case. Allows slightly greater local clarity. I am indebted to "Donald" for pointing out how PrtSc can capture such images to the clipboard for insertion into image handling software for subsequent cropping and enhancement. Otherwise I could not have shared these Google images of the G&J in the observation gallery.

The round object on the rear of the back plate is covered in cooling slots. Purpose still unknown. A plate rectifier perhaps to lower the mains to that required by the G&J WT movement?   



There is a dark "shadow" from a missing component on the face of the back plate just to the right of the Hipp Toggle on the pendulum. Where the paint has not been bleached by long exposure to the light there are drilled holes. It could have been a toggle damper. These are used on most Gents' WTs. It might have been something else.










Having browsed at length using various search terms, without much success, I  would be grateful for any images or links to other images of Gillett & Johnston Waiting Train, electrically driven, tower/turret clock movements. It seems unlikely that only two exist in the entire world.

My thanks go to Jim and Bart for generously sharing images and links.

Click on any image for an enlargement.
Back click to return to the text.

*

2 comments:

donald said...

If all else fails "print screen" copies the screen to the clip board so it can be pasted into any image processing programme, or paint or word.

But you knew that.

Chris.B said...

Hello Donald,

I think I must have missed that class. ;-)

Thank you,
Chris