Function Escapements and Drops

eri231

Registered User
Jan 13, 2012
1,850
910
113
torino italy
Country
Region
"The free travel of the escape wheel after impulse and before locking again"
Britten's Old Clocks & Watches
Regards enrico

Sorry but gmorse beat me
 

miguel angel cladera

Sponsor
Donor
Jul 29, 2019
478
956
93
51
Country
Hi Miguel,


The drops in any escapement are the period or distance between one tooth being released by one pallet and another tooth being stopped by the other pallet. There has to be some drop on each pallet, usually measured as a degree of rotation, but it should be minimal and both drops should be the same. In a verge, the drops are adjusted by moving the escape wheel inner pivot, (closest to the verge staff), laterally, since the verge flags are acting on diametrically opposite escape teeth. This is done in English watches by a simple dovetail slide on the potence, but on French watches mostly by a more complex arrangement of slides with screw adjusters.

Regards,

Graham
Thank you so much Graham! now I can understand.
 

John Matthews

NAWCC Member
Sep 22, 2015
4,080
2,187
113
France
Country
Region
I thought it was defined by the travel of the escape between release and locking, and that the definition applied to all escapements. Am I mistaken?

Edit (after checking Gazeley)

There is a subtle difference between the two definitions given by Graham & Enrico and I believe the difference will vary with escapement type.

John
 
Last edited:

gmorse

NAWCC Member
Jan 7, 2011
14,554
3,672
113
Breamore, Hampshire, UK
Country
Region
Hi Jeffery,
How are drops adjusted in levers?
If the pallets are jewelled in the modern way, ie, visibly, the stones may be moved relative to the pallet frame, but that isn't something to be fiddled with without knowing exactly what one is doing, because moving one affects the other. In English levers with hidden jewels, they can't easily be moved.

In cylinders the drops are defined by the dimensions of the escape teeth and the cylinder itself. In English duplexes, the situation is similar, it's governed by the escapement geometry which isn't normally alterable without rotating the rollers on the staff.

Regards,

Graham
 
  • Like
Reactions: Incroyable

John Matthews

NAWCC Member
Sep 22, 2015
4,080
2,187
113
France
Country
Region
Graham,

Does unlock and impulse occur at the same instance in all escapements?

John
 

gmorse

NAWCC Member
Jan 7, 2011
14,554
3,672
113
Breamore, Hampshire, UK
Country
Region
Hi John,
Does unlock and impulse occur at the same instance in all escapements?
No, and I don't think it's ever completely instantaneous. Since the escapement is continously starting and stopping the whole train, there's an inertia factor.

In a detached lever, the escape tooth slides along the locking plane until it reaches the locking corner, when it starts to move along the impulse plane, but in making that transition, it goes from the impulse pin on the balance moving the lever via the fork, (using the stored energy in the balance spring), to the opposite face of the fork snapping across to the pin, (now being driven by the mainspring via the escape wheel), and hence impulsing the balance. In order for this to work at all, there has to be some freedom between the pin and the fork, since the pin has to enter and exit the fork freely, which means that there's a short delay. The better made and designed the watch, the closer this tolerance can be, but there are also the necessary freedoms in the escape, lever and balance pivots. The effect of draw, in ensuring that the lever is always pushed into the bankings, will influence the depth of locking but that doesn't have any impact on this matter.

In a cylinder, there's no lever with a fork to introduce freedoms, but the freedoms are still there, in the necessary clearances of the escape teeth on both the internal and external diameters of the cylinder, plus of course the pivot freedoms as before. There aren't so many impacts involved here. The balance is only detached during the brief period of the drops.

Lastly, in a duplex, in which the balance is directly impulsed by a pallet on the roller, (as in a detent), there must be some clearance to allow the reverse swing, so the unlocking action of the long escape teeth with the small slotted roller on the staff has to be very slightly ahead of the short escape teeth which hit the impulse pallet. Being only impulsed in one direction, (again like a detent), it's almost detached on the reverse swing, except for the unlocking teeth on the tiny slotted roller.

Regards,

Graham
 

John Matthews

NAWCC Member
Sep 22, 2015
4,080
2,187
113
France
Country
Region
Thank-you Graham.

My point was that there is rotation of the escape as soon as it becomes unlocked, and therefore there is rotation of the escape after unlocking and before impulse. It follows that there is a difference in the amount of escape rotation measured between impulse and locking as opposed to between unlocking and locking. Is this not what is demonstrated by the three 'peaks' on a trace, unlocking, impulse, locking ...

1665157357601.png


John
 

Bernhard J.

NAWCC Member
Sponsor
Jan 10, 2022
1,032
1,085
113
Berlin, Germany
Country
Region
But are both processes not jointly called "drop", even though the drop is asymmetric, depending on the type of escapement and the particular escapement adjustments?
 

gmorse

NAWCC Member
Jan 7, 2011
14,554
3,672
113
Breamore, Hampshire, UK
Country
Region
Hi John,
My point was that there is rotation of the escape as soon as it becomes unlocked, and therefore there is rotation of the escape after unlocking and before impulse.


Yes, the escape wheel must begin to move after the tooth passes the locking corner, at which point the fork has to switch from being driven by the pin to driving it to deliver impulse. The delay at this point is miniscule but it is finite.

See what you make of the video.

Regards,

Graham
 

John Matthews

NAWCC Member
Sep 22, 2015
4,080
2,187
113
France
Country
Region
Graham - excellent.

Am I correct in assuming that the delay is greater with duplex & detent escapements?

John
 

gmorse

NAWCC Member
Jan 7, 2011
14,554
3,672
113
Breamore, Hampshire, UK
Country
Region
Hi John,
Am I correct in assuming that the delay is greater with duplex & detent escapements?
For cylinders, my feeling is that the delay would be shorter, since the escape wheel is engaging directly with the balance and not via an intermediary, although I have no way of proving this.

Lastly, in a duplex, in which the balance is directly impulsed by a pallet on the roller, (as in a detent), there must be some clearance to allow the reverse swing, so the unlocking action of the long escape teeth with the small slotted roller on the staff has to be very slightly ahead of the short escape teeth which hit the impulse pallet. Being only impulsed in one direction, (again like a detent), it's almost detached on the reverse swing, except for the unlocking teeth on the tiny slotted roller
I think it's a design issue here; if the angle of separation between the two rollers isn't precisely matched to the escape wheel dimensions, it could result in a longer delay.

Generally, any delay between unlocking and impulse results in lost impulse and hence reduced amplitude. Remember that the first noise in the trace is the impulse pin hitting the side of the fork, so it's the start of unlocking not its completion. All this makes George Savage's lever escapement seem more impressive, since his design separated unlocking from impulse.

Regards,

Graham
 
  • Like
Reactions: John Matthews

Dr. Jon

Moderator
NAWCC Member
Dec 14, 2001
7,466
1,866
113
New Hampshire
Country
Region
My 2 cents.

There is no significant delay in most escapement between unlock and impulse.
On the lever the escape tooth moves from the locking to the impulse side of the pallet. The feature called the unlock is actually the sound of the impulse pin hitting the fork to begin the unlock.

On impulse the second sound is largely due to the impulse pin moving from one side of the fork to the other and striking it.

To me, the drop on a lever is the movement from impulse to lock.
 

gmorse

NAWCC Member
Jan 7, 2011
14,554
3,672
113
Breamore, Hampshire, UK
Country
Region
Hi John,

By the time the impulse has been given, the balance is moving away and the lever has moved into position ready to lock the next escape tooth. As it locks, the draw pulls the lever into the banking, ready for the next unlocking event, which ensures that it always starts from a predictable position and hence in a consistent time. As Jon says, we're talking about very small time intervals here.

Regards,

Graham
 

John Matthews

NAWCC Member
Sep 22, 2015
4,080
2,187
113
France
Country
Region
All I am trying to establish is the correct definition - whether the difference is small or not.

This thread concerns a verge. Two definitions were given in terms of the rotation (period or displacement) of the escape wheel. I simply wish to know which is correct as it seems there is a difference, albeit small.

John
 

gmorse

NAWCC Member
Jan 7, 2011
14,554
3,672
113
Breamore, Hampshire, UK
Country
Region
Hi John,
All I am trying to establish is the correct definition - whether the difference is small or not.
I think much of this discussion revolves around exactly when the 'impulse' happens, because this has a finite duration, and the size of the drop depends on which point it's measured from;
  1. is it immediately on unlocking when the tooth starts its travel along the impulse plane,
  2. when the lever fork hits the opposite side of the impulse pin,
  3. when the tooth drops off the impulse plane,
  4. when the other pallet drops into the path of a tooth,
  5. when the draw on the pallet pulls it into the banking?
The Britten's definition refers to the 'free travel' of the escape wheel before locking, so on that basis, my option 3 would apply, because after unlocking, the escape tooth is still in contact with the pallet until it drops off the impulse plane. Its 'free travel' ceases as soon as a tooth touches the other pallet's locking plane.

My more long-winded definition was less precise, but on reflection, I think this interpretation of the Britten's phrase will do well enough for all practical purposes.

PS: Jon, I think this discussion would be better in a thread of its own, and Miguel and John M both agree I believe. It went awry about post #33.

Regards,

Graham
 

John Matthews

NAWCC Member
Sep 22, 2015
4,080
2,187
113
France
Country
Region
revolves around exactly when the 'impulse' happens
Graham surely if the definition is in terms of rotation of the escape, the key events are the release of the escape so that it can rotate and the point when the rotation ceases. The impulse will occur at some point after the escape is released - it cannot happen while the escape is locked.

By introducing the term 'free travel' is Britten restricting the rotation of the escape when there is no contact with the pallets?

Does this mean we have
  1. commencement of rotation
  2. commencement of free travel
  3. impulse
  4. end of free travel
  5. locking
So that Britten's definition is the amount of rotation of the escape between 2 & 4.

John
 

Dr. Jon

Moderator
NAWCC Member
Dec 14, 2001
7,466
1,866
113
New Hampshire
Country
Region
A additions.

I think the original issue is definition of the drop and this is movement from impulse to re-lock.
My view is that the change from unlock to impulse is virtually instant for levers, duplexes, and cylinders. This is because the escape wheel tooth stays in contact with the roller from unlock to impulse.

In detent escapements, the unlock and impulse are on different surfaces and the impulse may or may not be after unlock.

In an failed effort to write an article for the NAWCC Journal I collected "sonograms" of various escapements.

Here is a verge
1665245392583.png

Verge


1665246023770.png


Cylinder

1665245956538.png

Marine spring detent chronometer The passing is barely visible.

1665246130819.png

Pivoted Detent

The small "blip is the passing.

This is a digitized Lever

1665246322422.png

All escapements must unlock, impulse, and re-lock at least once per cycle.
 
Last edited:

gmorse

NAWCC Member
Jan 7, 2011
14,554
3,672
113
Breamore, Hampshire, UK
Country
Region
Hi John,
By introducing the term 'free travel' is Britten restricting the rotation of the escape when there is no contact with the pallets?
He's saying that its free travel can only occur when there's no contact with the pallets.

The sequence in detail is this:
  1. The lever is locked on an escape tooth and is on the banking.
  2. The balance is on its return swing and the impulse pin enters the lever fork and contacts one side of it. The fork is driven by the impulse pin. This is the first noise peak.
  3. The fork starts to move the locking plane along the locked tooth towards the unlocking corner. In doing so it has to cause recoil because of draw, which removes energy from the balance.
  4. The locked tooth reaches the unlocking corner.
  5. The tooth, now unlocked, starts to move along the impulse plane which causes the opposite side of the fork to contact the pin and become the driver of the impulse pin. This is the start of the impulse and is the second noise peak.
  6. The tooth reaches the end of the impulse plane. End of impulse.
  7. The balance continues to swing and the impulse pin leaves the fork.
  8. The other pallet is now in position to lock another tooth.
  9. The tooth hits the locking plane of the other pallet and draw moves it further into the banking. This is the third noise peak.
  10. Da capo for the other pallet on the reverse swing.
'Free travel' of the escape wheel begins at the end of 6 and ends at 9, which constitutes the 'drop'.

Regards,

Graham
 

John Matthews

NAWCC Member
Sep 22, 2015
4,080
2,187
113
France
Country
Region
Graham

This is for the lever, is it not? I think the definition should be applicable to all escapements and I do not believe this sequence can be applied to duplex or detent. Perhaps my understanding is in error. I thought free travel begins before the impulse in these two escapements.

John
 

gmorse

NAWCC Member
Jan 7, 2011
14,554
3,672
113
Breamore, Hampshire, UK
Country
Region
Hi John,
This is for the lever, is it not? I think the definition should be applicable to all escapements and I do not believe this sequence can be applied to duplex or detent.
...
I thought free travel begins before the impulse in these two escapements.
Yes, it is for the detached lever. I think you're right about the duplex and detent escapements, but although they must have drops, I'm not sure that it's possible to arrive at a generic definition. The duplex has a very tiny amount of free travel between unlocking and impulse, but after impulse the next locking tooth is moving until it's stopped by the ruby cylinder, which is also free movement.

The detent escape wheel is free between impulse and the next tooth locking.

Impulse in both these escapements is given by a blow rather than a steady push, the sequences for both are rather simpler than that for the lever.

Regards,

Graham
 

John Matthews

NAWCC Member
Sep 22, 2015
4,080
2,187
113
France
Country
Region
Graham

I just feel that defining it as simply the free movement of escape is an adequate generic definition. Involving the impulse I think makes it more complicated and is not necessary.

John
 

gmorse

NAWCC Member
Jan 7, 2011
14,554
3,672
113
Breamore, Hampshire, UK
Country
Region
Hi John,

I think generic definitions are indeed adequate for some purposes, but sometimes it's necessary to have a more detailed understanding of a mechanism, especially one as deceptively simple as the lever escapement. I'm sure old Thomas Mudge never appreciated what he'd started!

Regards,

Graham
 

Forum statistics

Threads
177,601
Messages
1,556,545
Members
53,629
Latest member
nicobirdie
Encyclopedia Pages
909
Total wiki contributions
3,058
Last edit
Watch Inspectors by Kent