1810~ Robert Roskell '3 wheel rack lever discovery'

[John Matthews]

I purchased this Roskell rack, which I believe dates from ~1810, at the beginning of the month. This thread is initially a photographic record of what I discovered as I dismantled it. In subsequent posts I will record my attempts to service it and any problems I encounter. As I have expressed previously, for me, the understanding of the mechanism is as important as performing the service. I will now add to that the interest of what it reveals regarding the process of manufacture and the history & method of repair. I have many photographs and so the thread will open with multiple posts.

The immediate attraction of the watch was the perfect one piece cream dial with a 15 second subsidiary and that it was in 'working order'. Superficially, on arrival, it appeared to be a clean state and I sprang into life immediately when I gave it a single turn wind. As can be seen the only the minute hand is in good condition. (I fitted the second when I tested it). The top plate has the typical Lancashire features of the early C19th, plain steel balance, nicely engraved balance cock and characteristic regulation scale markings.



Removal of the dial revealed the underside of the pillar plate devoid of any movement makers identification marks but with the exposed parts in a clean condition, with no evidence of significant ill-treatment or corrosion. At this stage I did begin to see some evidence of possible reworking of the pivot holes.



The underside of the balance cock has a pair of shallow punch marks on both sections, which I believe are batch marks. I had thought that these cocks were machined from a single piece of brass, but the batch marks on both sections, has caused me to examine the step under the microscope. Although very thin, I do believe I can see evidence for a joint with the foot having the step to which the flat table has been attached. I would be grateful for some guidance on this observation. At a later stage the movement number has also been hand scored on the foot



The hairspring is very slightly coned but otherwise in reasonable condition. The pivots look satisfactory, but the pinion shows some worrying signs of wear. The index is in good clean condition.



To be continued ...

 

[John Matthews]

The underside of the bridge that houses the third wheel and escape pivots, reveals a number of signs of modification/re-working. The pivot hole for the escape is very dirty and appears to have been moved and re-plugged (not particularly neatly). Material has been removed at the opposite end of the bridge. The scribe marks defining the position of the third wheel pivot can also be seen. I did expect to find a pair of batch marks, but none were found. I believe there are two possibilities. Firstly the reworking may have obliterated them, secondly this is a replacement bridge.



The third wheel appears to me to be in good condition. The escape wheel teeth are generally good, although the tip of one or two do show signs of minor damage. The escape pinion has some wear and I think the shaft may have been drilled and the extended pivot supporting the second hand, replaced.



to be continued ..

 

[John Matthews]

Three of the rack teeth have been replaced. The repair is definitely functional, but clearly not a pretty sight!. At some point there has clearly been an altercation between the rack and the balance pinion! I'm sure Graham will provide an insight!



to be continued

 

[John Matthews]

The underside of the barrel bridge carries two distinct set of batch marks. A pair of marks marching those on the balance cock and four slightly large circular marks The fusee, barrel and centre wheel appear to be in good condition.



to be continued

 

[John Matthews]

The underside of the top plate shows further evidence of re-plugging of the pivot hole corresponding to that seen earlier on the bridge and the broken lever of the hack feature can be seen. The colour of the slide, housing the lower balance cap caught my eye – would this be bronze rather than copper?



Finally the underside of both plates contain two sets of batch marks. A pair of shallow punch marks as found on the balance cock and the barrel bridge with five rather large well shaped marks – I believe these are later and have been formed by a drill rather than a punch. It seems probable that there was originally five such marks on the barrel bridge, but one may have been lost when the chamfer was formed on the adjacent hole.



More as I progress to cleaning and re-assembly.

John

 

[John Matthews]

Two further photographs with the escape and rack mounted on the top plate, showing the rack in its two extreme positions against the banking pins.



John

 

[gmorse]

Hi John,

The 'extra' holes next to the escape wheel pivot in the pillar plate bar and next to the third wheel pivot in the top plate are intended as sight holes, to enable the meshing of the adjacent pinions to be examined more easily. They appear to be characteristic of Roskell's work. However, that escape wheel pivot hole has certainly been re-bushed in a different position slightly further away from the third wheel. It could be the result of a replaced third wheel, but less likely to be a planting error. It's nice to see the planting arcs for the third wheel on the underside of the bar, which would have been scribed with a depthing tool.

I had thought that these cocks were machined from a single piece of brass, but the batch marks on both sections, has caused me to examine the step under the microscope. Although very thin, I do believe I can see evidence for a joint with the foot having the step to which the flat table has been attached. I would be grateful for some guidance on this observation.

I think this is most unlikely unless it's been repaired, these were made from one piece of brass.

The colour of the slide, housing the lower balance cap caught my eye – would this be bronze rather than copper?

It's probably brass, but not gilt like the potence itself. As it's carrying an endstone, there would be little reason to make it of anything else.

Regards,

Graham

 

[gmorse]

Hi John,

Two further photographs with the escape and rack mounted on the top plate, showing the rack in its two extreme positions against the banking pins.

Thanks, these are exactly what I needed. With the rack in these two extreme positions, is it still fully engaged with the pinion? They also show what appear to be a couple of replaced teeth in the rack; a delicate job! They could explain why the pinion is so badly cut.

Regards,

Graham

 

[gmorse]

Hi John,

They could explain why the pinion is so badly cut.

I hadn't seen those pictures of the rack when I wrote this. Those reinforced, wider teeth may have been the cause of the pinion wear, but on the other hand they could equally have been the attempt to remedy it, because the original teeth all seem to be quite worn, just those central three are the worst.

Regards,

Graham

 

[John Matthews]

The 'extra' holes next to the escape wheel pivot in the pillar plate bar and next to the third wheel pivot in the top plate are intended as sight holes

As described by Graham ....

It's probably brass, but not gilt like the potence itself. As it's carrying an endstone, there would be little reason to make it of anything else

Graham - should be resolved when I clean it ..

is it still fully engaged with the pinion

Graham - I didn't check - I assume the best approach would be to mount the rack lever & escape between the plates and then position the balance and check the rack in the two extreme positions - is this the best approach?

John

 

[gmorse]

Hi John,

...I assume the best approach would be to mount the rack lever & escape between the plates and then position the balance and check the rack in the two extreme positions - is this the best approach?

Yes, that's the only way to be sure; I don't see how it could ever become disengaged if the banking pins are correctly planted, as they appear to be.

Regards,

Graham

 

[John Matthews]

Having mounted the escape & lever between the plates and put the balance in position, I very gently incrementally rotated the balance until it reached the end of permissible travel in both clockwise and anticlockwise directions. There was a definite recoil from the extreme positions, which I believe is when the lever engaged with the respective banking pin. The second photograph below has the lever banked against the inclined banking pin in the foreground at the end of the anticlockwise rotation of the balance; in the third photograph the lever is banked against the short banking pin (above the escape in the rear of the photograph) at the end of clockwise rotation of the balance.



I then removed the balance. The following photographs show the position of the rack looking vertically down towards the lower pivot hole for the balance. The first photograph corresponding to the lever banking against the inclined pin (second photograph above), the second photograph when the lever banked against the short pin.



Although there was no tendency for the balance shaft pinion to escape from the rack, it is clear that the rack is closer to 'approaching escape' when the balance is at the extreme of its clockwise oscillation (the second photograph above), i.e. when it engages with the short pin. In the photograph below it can be seen that the short pin (the one on the right) is vertical and shows no evidence of having been 'adjusted'. Whereas the longer pin has been bent in a manner that would allow very slightly more anticlockwise rotation.



When I was taking these photographs, my attention was drawn to the broken hack lever. I know that these are often deliberately disabled, often by shortening the lever, as appears to be the case with this example. Does anyone have a good photograph of a working example on a rack lever, so that I can they operate?



John

 

[Allan C. Purcell]

Hi john- The watch below I think is like yours.(Stops the anker) I also looked at the three Hornby Racks I have, and one like yours had been removed, (all of it) the second did not have a Hack, and the thired is totally different in that it stops the fusee, most unusual.Allan

 

[John Matthews]

Thanks Allan - that's helpful.

If you have the watch to hand, when you operate the hack can you describe how the horizontal portion of the brass wire moves and where exactly it makes contact with the lever?

John

 

[Allan C. Purcell]

You push the the little lever to the left and the arm flicks to the left onto the rear of the the flat part of the anker. See below, I bought it as a movement-then found the case for it-the marks I will leave for you-I know you like to do that-we all do.
The hands I put on too. Best Allan.

 

[Allan C. Purcell]

Hi John-some more Rack Levers with a Hack.
This is Number 23379, it is the same as yours, and operates has above.

This is again the sam number 8932-though this time-like yours filed off.


This is an Early Rack Lever by M.I.Tobios & Co.2534 no hack but did have a balance break. No slot a round hole. (re-moved).

Hope these help John-I think this style of hack was the one most used I have it on other watches-but they are not Racks.

Best wishes,

Allan.

 

[Keith R...]

Thanks John, I need to check my number with Allan's most recent data base. Someone asked
and I said 1812. It now looks to be earlier.

PS...Hairspring is fine, keeps perfect time.

Keith R...

 

[gmorse]

Hi John,

...when you operate the hack can you describe how the horizontal portion of the brass wire moves and where exactly it makes contact with the lever?

The horizontal part of the brass pin contacts the back of the pallets at a shallow angle to stop the escapement. There's enough spring in the brass wire to prevent any damage.



Regards,

Graham

 

[John Matthews]

The horizontal part of the brass pin contacts the back of the pallets at a shallow angle to stop the escapement. There's enough spring in the brass wire to prevent any damage.

Graham - thanks for the photograph and for identifying exactly which part of the lever it acts upon

John

 

[gmorse]

Hi John,

I expect this is a more useful image of the stop lever than just the lever by itself and shows it in the stop position. The lever actually engages with the back of the entry pallet. Don't take too much notice of the rack, it's still a work in progress.



Regards,

Graham

 

[John Matthews]

Encouraged by the photographs provided by the recent contributors, I decided to repair the hack feature by replacing the broken/cut brass wire; I have reached a point where I need some guidance.

To remove the old wire from the lever, I cut it off as close as possible to the lever and filed the remainder flush. I mounted the lever in my small hand clamp, placed it on the vice and punched the stub out of the lever. I then threaded a long tapered pin into the vacant hole and tapped it gently to secure it.



With the lever mounted between the plates, I rotated the lever so that it was banked against the pin closest to the hack slot, I secured the lever in that position using some blue tac.



By carefully removing the top plate, without disturbing the position of the lever, I was then able to determine the height above the pillar plate of the back of the pallets and by holding the pin with a pair of round nose pilers, bend the pin horizontal at that height above the plate.



This is the point where I need advice ...

My thought is that I need to re-position the horizontal wire so that is directly over the lever pivot hole, when the lever is in the position shown in the second photograph above and then reduce the length so that it is just short of the back of the pivot. If my logic is correct, with the wire so positioned, the lever will be able to oscillate freely. When the wire is moved to the position shown in the third photograph, it will engage with the back of the pallets towards the edge of the plate, as the lever approaches the banking pin furthest from the hack slot.

Is my logic correct?

John

EDIT - Graham - you posted after I started - it looks as if my logic is almost correct

 

[John Matthews]

Graham - my thanks

I expect this is a more useful image of the stop lever

Your photograph Is exactly what I needed to work from ...

John

 

[John Matthews]

I now have a functioning hack feature as the photographs illustrate.

The brass wire engages the rear of the pallets as Graham's photograph in post #20. The oscillation of the lever is stopped before it is able to reach the short banking pin, i.e. stopping the anticlockwise oscillation of the balance and the clockwise oscillation of the lever, when viewed from the top plate.

Hack preventing oscillation - the third photograph shows the lever stopped before reaching the small banking pin ...



Escapement working normally - the second photograph shows the lever against the small banking pin ...



the top plate removed provides a clearer view of the two positions of the brass wire ...



I just need to tidy the ends of the hack wire before starting to clean and lubricate the movement.

John

 

[gmorse]

Hi John,

When you come to secure the brass wire in the lever, you need to make sure that it can't rotate, so a close look at the hole in the steel may reveal a chamfer or perhaps some grooves to make it more solid when it's riveted in.

Regards,

Graham

 

[John Matthews]

Graham - I have not seen any evidence of features on the lever, to assist securing the pin.

I used a tapered pin and when I originally bent it, I tested it and it was very difficult to rotate. Now in its final position, having marked the lever and wire so that I could tell if the pin moved, I have removed the excess length, filed it almost flush and then gently tapped it flat with the small hammer. I then used a sharp punch to indent the centre in an attempt to secure it further by expanding the end. The pin does not appear to have changed its position and had no tendency to move, when I applied pressure to the bent portion of the pin.

John

 

[gmorse]

Hi John,

...I then used a sharp punch to indent the centre in an attempt to secure it further by expanding the end. The pin does not appear to have changed its position and had no tendency to move, when I applied pressure to the bent portion of the pin...

That should be tight enough now; the technique is correct, especially if there's no shoulder on the other side.

Regards,

Graham

 

[John Matthews]

I have now finished servicing Roskell rack and it has successfully run to the end of a complete wind.

Now I need to come clean and say what I omitted from the service (sorry Graham!). I did not dismantle the fusee, I simply cleaned the outside by hand, neither did I remove the spring from the barrel. I did remove the cap to perform a visual inspection and lubricated the spring will a small amount of Moebius 8200. My rationale is that the movement will only be run a few times a year and, prior to dismantling, the winding was smooth and I could hear the internal fusee clicks operating during the wind.



The hack feature is working as it should and I have fitted the most appropriate hands that I have in my limited spares. I appreciate the main hands do not match and both the hour hand and the second hand are short. I am currently checking the time-keeping. Part way, it looks as if t may be a touch slow but not excessively so.



John

 

[gmorse]

Hi John,

...Now I need to come clean and say what I omitted from the service (sorry Graham!). I did not dismantle the fusee, I simply cleaned the outside by hand, neither did I remove the spring from the barrel. I did remove the cap to perform a visual inspection and lubricated the spring will a small amount of Moebius 8200. My rationale is that the movement will only be run a few times a year and, prior to dismantling, the winding was smooth and I could hear the internal fusee clicks operating during the wind...

That's a good result, and I think you can be excused about the fusee and barrel, (just this once mind!); at least with Harrison's maintaining power there are two clicks, and it will function at a pinch with just one. The mainspring looks as though it fits the barrel well, applying the rule of thirds to the barrel radius rather than its area, which is what would have been used by the makers in all probability. Using areas results in a slightly shorter spring with a little more room to develop, but it's marginal.

Well done!

Regards,

Graham

 

[Keith R...]

Pretty darn great John! One week on a Roskell Rack Lever from 1810.

I'm still waiting for the Lawrance Rack from 1816 to come home, acting
somewhat like a Rack in all positions.

My hat's off to you and the coach!

Keith R...

 

[John Matthews]

Graham/Keith - my thanks for your kind comments. Keith I fully endorse

and the coach!

The movement has now been running face up for 4 hours and it has maintained time exactly aligned with the computer as far as I can tell, certainly within 15 seconds - pretty darn good for a 200 year old rack ...

John

 

[John Matthews]

The movement has now worked through a 24 hour cycle and over that time it has gained three minutes. I will make a small adjustment and run it for a further 24 hours and then return it to its place in my cabinet.

John

 

[John Matthews]

To close ... made two further adjustments and in the last 25 hours it lost, at most, 3 seconds - its now having a well earned rest in the cabinet.

John

 

[Keith R...]

John, great job and it's a pleasure to follow along with this old watch! I love the cream
dial as I prefer them most. It might just be me, but many of my Liverpool watches sport
original cream dials, more so than my London examples. Now that could be by personal
choice, I'll have to do a bit of study and see.

Keith R...