Wales and Marches Horological Society :
For people interested in horology: clocks, watches, scientific instruments, barometers or just the passing of time!
Repairing a Daniel Quare Spring Clock
Lessons learned – Beware of assumptions
Dennis Radage
A Daniel Quare spring clock was recently brought to my workshop for examination and possible repair since it would not run, more correctly, when started, it slowed and then stopped within a minute or two.
The clock, signed on the dial, chapter ring and backplate; ‘Dan. Quare London’, can be seen in Figures 1 & 2. This is a time and strike, pull quarter repeat spring clock in an ebony-veneered case with a gilt-decorated dome top. The case has open fret to the top front door rail allowing the strike and quarter repeat to be clearly heard.
fig 1 pull quarter repeat spring clock
fig 2 decoratively engraved and signed backplate
The clock has a rectangular dial, a date feature above the Roman VI-hour position, a false (mock) pendulum and two subsidiary dials at the top; the right being for strike/silent and the left for rating, lifting or lowering the pendulum by a rack and pinion mechanism on the backplate. The clock dates to circa 1705.
Figure 2 shows the decoratively engraved and signed backplate, the short pendulum and the engraved cover of the rack and pinon rating mechanism.
Several attempts were made to start the clock, but it would slow and then stop every time after a minute or two. Obviously there was an issue.
The problem could not be identified without a detailed examination, which meant removing the movement from the case. The movement is held in with two brackets screwed to the backplate and case sides.
fig 3 the movement removed from the case for examination
fig 4 The complex mechanisms on the front plate of the movement
Figure 3 shows the movement out of the case. Note that the pull quarter cord had to be disconnected before the movement could be separated from the case. The hour bell and quarter bells can be seen to the top of the movement. Visual examination could not identify any issues, the clock continued to slow and stop. As with just about all spring clocks of this period, the going train had a verge and crownwheel escapement.
The hands were removed, then the dial was separated from the movement, this of course disconnected the subsidiary dials from their arbors and levers.
Figure 4 shows the movement front plate. The motion work, false pendulum, rack striking mechanism and the quarter repeat mechanism are all mounted on the front plate along with the mainspring set-up clicks and ratchets. The minute hand was re-installed so as to observe hour and quarter striking and repeating. Careful examination did not identify any issues or fouling parts. Note that this is reasonably complicated clock.
The clock would strike correctly and the pull-quarter repeat also functioned appropriately, no issues were found on the strike side. The issue was deemed to be associated with the going train.
Some history; this clock was sent to the UK a few years back to have some restoration-work performed. It is known that the clock was fully dismantled at that time, cleaned and professionally restored. As can be seen in Figures 1 through 4, the case and movement are clean and visually well restored. The clock performed all functions correctly when returned to its Canadian owners. Given this history, I was not expecting to find any serious faults.
The hour bell was removed, then power to the various wheels was checked. There was ample power to the contrate wheel which drives the pinion of the crownwheel. It seemed clear that there were no issues with the mainspring, fusee/great wheel, centre arbor with its wheel or the contrate wheel. It was observed however, that the beat was inconsistently slightly uneven.
The next task was to let down the mainspring of the going train. This allowed the side shake (freedom of pivots in their holes) to be checked. No issues were found here.
The pendulum, then back cock were removed followed by the verge shaft and pallets. These were examined and found to be in good order.
fig 6 Bottom potence and contrate wheel clearly visible
fig 5 View showing the crownwheel with the top and bottom potence
The crownwheel top potence (pivot) (as can be seen in Figure 5) was removed. This now allowed the removal of the crownwheel with its arbor and pinion that interface with the contrate wheel. The crownwheel was rotated that identified a slight unevenness. This was leveled. As no other issues were observed, the movement was reassembled, replacing all parts including the back cock and pendulum. The going train mainspring was wound slightly, the contrate wheel again having sufficient power, but although the beat was now quite even, the movement still would not run for more than a minute. It was noticed that there was quite a large drop of the pallets onto the crownwheel teeth.
The mainspring was again let down, the pendulum and back cock removed, the pallet arbor was removed as was the top crownwheel potence, allowing the crownwheel to be lifted out yet again. The crownwheel arbor is pivoted in a lower potence which can be seen in Figures 5 and 6. It was noticed that the lower potence was blackened around the pivot hole. The lower potence was removed for examination. Both the upper and lower potence parts have two locating pins and each are held to the inside of the backplate by a single screw. Note that each time the going train is dismantled, the mainspring must first be let down. After reassembly, the mainspring must again be wound sufficiently to power the train. This winding was achieved by turning the arbor of the mainspring with a let-down key, the ratchet and click holding the slightly wound position.
Examination of the lower potence exhibited blackened and gummed lubrication making this pivot area quite dirty, Figure 7. The lower pivot of the crownwheel arbor rests on a hardened screw which is mounted into the bottom of the potence, directly under the pivot hole. This acts in a similar manner to that of an endstone. The screw is used to raise and lower the crownwheel arbor, thus crownwheel, which allows adjustment of the drop of the verge pallets onto the crownwheel teeth.
The potence was cleaned, then examined. This looked to be in good condition with no serious wear. The crownwheel bottom pivot was also cleaned and then checked for fit in the lower potence pivot hole. A small amount of fresh lubrication was applied. The adjusting screw was turned one-half of one turn so as to improve the pallet drop. Since no additional issues were noticed, the assumption was that this assembly was now good.
As previously indicated, I was not anticipating serious issues because of the previous restauration work. Sorry to admit that this wrong assumption led me to reassemble the clock again and power the mainspring. The assembly was now clean and the pallet drop had been improved. The clock was set in motion, with slightly improved performance. The clock was trying to function, but the problem had obviously not been corrected, even though I now had several minutes of running before the clock slowed and stopped.
fig7 Bottom potence removed. Crownwheel pivot blackened.
fig8 The bottom potence showing the correct shape of the pivot, and the positioning of the hardened steel adjusting screw
Instinct told me that the problem was indeed related to the crownwheel assembly, but there was nothing apparently wrong, at least, not obvious. It was time to take a break and read a little about early verge escapements, I read Practical Clock Escapements by Laurie Penman. I focused on the chapter describing verge pivots. One image in this book struck me as being quite different from what I could recall, related to the shape of the lower crownwheel arbor pivot. Figure 8 is an image similar to page 29, Figure 2.29 of the Penman book. This suggests that the correct shape of the pivot should have a rounded end, thus having just a single contact point.
The mainspring was again let down and the pendulum, back cock, verge arbor and top potence were removed again. The crownwheel was removed. What became very clear this time is shown in Figure 9. The bottom pivot end was flat with a sharp edge. This pivot rotates on the hardened steel adjusting screw. A flat pivot end would introduce considerable friction when rotating on this screw. Using stones, this pivot was properly rounded, then polished. After cleaning, the clock was again reassembled and power was applied to the mainspring. This time, when set in motion, there was a marked improvement. The pendulum swung with vigour, there was no stopping it this time. Problem identified and corrected.
fig9 Bottom crownwheel arbor pivot has a flat bottom with sharp edge. This was causing considerable friction, wear and blackening of the pivot.
Since the mainspring had been let down several times, with the gut line still half wound on the fusee, the mainspring was given power several times until all of the line was wound off of the fusee. This took several days. At this point the clock is in its unwound state. These few running days allowed continued observation of the pivot and further assurance that the problem had indeed been corrected.
The going train mainspring was now set with just enough power so that the clock would run if there was more line.
The clock was now wound with a key to the fusee winding arbor until fully wound. The movement was allowed to run for a week before the dial was replaced, connecting the subsidiary dials in the process as well as replacing the hands. The false pendulum was positioned such that it swung freely and the star wheel and snail were synched to the position of the hour hand. The hour bell was replaced, then the movement was returned to its case and locked in position with mounting brackets. The pull quarter cord was reconnected to complete the project.
It is quite surprising how the shape of the lower crownwheel pivot would make such a difference. Why was it flat instead if rounded? Likely we will never know.
One should never assume, not even if the clock has been recently restored. I am only sorry that it took several attempts to isolate the issue. The evidence was there, but I initially missed it. Unfortunately, whether the movement is dismantled once, or several times, the service charge remains essentially the same. Lesson learnt, this kind of problem will not be missed again.