The Impulse Roller & Shape of Pallet


Michael Gilligan
 
Edited

As a new member, I thought I should start by contributing something

I have posted a copy of Fig. 101 from ‘Electric Clocks’ and my plot of the function.
 https://groups.io/g/synchronome1/album?id=256625

Hoping this is of interest

MichaelG.


Norman Heckenberg
 

Hi Michael,

That is an interesting contribution. Can you plot the pallet shape that corresponds? People often want to make an impulse pallet and worry about how to get the shape right.

Norman

 

From: synchronome1@groups.io <synchronome1@groups.io> On Behalf Of Michael Gilligan via groups.io
Sent: Saturday, 14 November 2020 10:42 AM
To: synchronome1@groups.io
Subject: [synchronomeelectricclock] The Impulse Roller & Shape of Pallet

 

[Edited Message Follows]

As a new member, I thought I should start by contributing something

I have posted a copy of Fig. 101 from ‘Electric Clocks’ and my plot of the function.
 https://groups.io/g/synchronome1/album?id=256625

Hoping this is of interest

MichaelG.


Chris
 

Hello Michael and Welcome to the forum.

Let us not forget that the impulse curve is also rotated about radii struck from the center of the rod. Which makes me wonder how the factory shaped the impulse curve on the raw pallet casting. Perhaps they used a high speed, copy lathe using a pantograph mechanism?

My own [pre-gravity arm buffer] 1919/20 Synchronome had apparently been run for decades with the pallet badly skewed to the plane of the arc. Requiring careful work to reshape the impulse curve. I used the highest speed in my old S&B lathe and various grades of abrasive paper. Wrapped and glued tightly around a slightly undersized rod to reshape the impulse curve in 3-dimensions. It would have required much higher speed to allow a form tool to shape the impulse curve due to extreme chatter and the weakness of the original casting.

Only after re-assembly did I realise that I had altered the overall length of the pallet slightly. Requiring the pendulum be moved bodily to the left. With further adjustments to suit. It has been running steadily for at least 35 years in my care so my "botching" seems to have worked. Or perhaps Synchronomes can shrug off the worst offences that can be thrown at them. Like being run out of line for long enough to wear the original pallet curve askew. New pallet castings became available later but it would been even more difficult to copy the original exactly than to "repair" it.

Regards

Chris.B


Michael Gilligan
 

Thanks for the welcome, Norman [and Chris]

The ‘Electric Clocks’ book is available at archive.org in various formats
https://archive.org/details/in.ernet.dli.2015.177424/page/n1/mode/2up
... and I find it convenient to have the ‘pdf with text’ on my iPad

Your question, Norman, is actually what prompted me to do the plot ... it was my way of satisfying myself that I understood the text on pp179-180
Hope-Jones, having prescribed what impulse he wanted [my green line] tells us that Shortt provided the mathematically pure function that would produce it [my red line].

Clearly, the relationship is ‘reversible’ ... so a cutter having the diameter of the proposed roller can be driven via the ’bump’ to machine the requisite profile on the pallet.
I have not yet found any corroborating evidence, but I presume that this would have been done on a Taylor Hobson, or similar, Pantograph.
Note: When I worked at KODAK, such things were done very successfully in the ‘development workshop’ using a hardboard template and a reduction ratio of 10:1

MichaelG.
.

P.S. __ I am adding another plot, showing a ‘zoomed-out’ view of the two functions


Ian Richardson
 

Chris,

Interesting experience with reshaping your impulse pallet.  Some years ago I actually machined one from a solid lump of brass (and it's still running well).  Until then, I wasn't aware of the radius on the impulse face.  To machine this, I mounted the pallet on rod clamped to a moveable, but essentially fixed, vertical slide and ran a milling cutter in the lathe chuck.  You will have to imagine the set-up.  By this means, I could manipulate the pallet as I milled the impulse face.  

Having seen many Synchronomes over the years, I think that there IS a great deal of in-built tolerance to the design which is a testament to the original idea.  Note, however, that very similar results can be obtained with a straight, flat impulse pallet - as Mr Parsons & Mr Ball found.

Cheers,
Ian R



-----Original Message-----
From: Chris <chris.b@...>
To: synchronome1@groups.io
Sent: Sat, 14 Nov 2020 7:24
Subject: Re: [synchronomeelectricclock] The Impulse Roller & Shape of Pallet

Hello Michael and Welcome to the forum.

Let us not forget that the impulse curve is also rotated about radii struck from the center of the rod. Which makes me wonder how the factory shaped the impulse curve on the raw pallet casting. Perhaps they used a high speed, copy lathe using a pantograph mechanism?
My own [pre-gravity arm buffer] 1919/20 Synchronome had apparently been run for decades with the pallet badly skewed to the plane of the arc. Requiring careful work to reshape the impulse curve. I used the highest speed in my old S&B lathe and various grades of abrasive paper. Wrapped and glued tightly around a slightly undersized rod to reshape the impulse curve in 3-dimensions. It would have required much higher speed to allow a form tool to shape the impulse curve due to extreme chatter and the weakness of the original casting.

Only after re-assembly did I realise that I had altered the overall length of the pallet slightly. Requiring the pendulum be moved bodily to the left. With further adjustments to suit. It has been running steadily for at least 35 years in my care so my "botching" seems to have worked. Or perhaps Synchronomes can shrug off the worst offences that can be thrown at them. Like being run out of line for long enough to wear the original pallet curve askew. New pallet castings became available later but it would been even more difficult to copy the original exactly than to "repair" it.

Regards
Chris.B


Bepi
 

Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
--
Bepi


H Hal
 

hi Bepi,


can you explain the difference between phase and time in relation to impulse or are you using the terms interchangeably?


regards


hal

On 14/11/2020 12:42, Bepi wrote:
Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
--
Bepi


Chris
 

Ian,

Thanks. I hadn't thought of using the vertical slide and a milling bit at the time.
I suppose I could have silver soldered a new "nose" section onto the original pallet casting and worked the impulse curve on that.
The earliest pallets were of a different design to later ones. So I wanted to keep the original for authenticity.
It's odd to think that my "False Pediment" Synchronome has now become a true, 100 year old antique. Yet is still functional.

The Gents WT uses an inverted impulse pallet design to great effect. Relying only on a sharply curved shoulder and a rising roller.
Though repeatable accuracy of impulse force is not needed in this case. Since it is not a timekeeper in its own right.

Regards
Chris.B


On 14/11/2020 10:38, Ian Richardson via groups.io wrote:
Chris,

Interesting experience with reshaping your impulse pallet.  Some years ago I actually machined one from a solid lump of brass (and it's still running well).  Until then, I wasn't aware of the radius on the impulse face.  To machine this, I mounted the pallet on rod clamped to a moveable, but essentially fixed, vertical slide and ran a milling cutter in the lathe chuck.  You will have to imagine the set-up.  By this means, I could manipulate the pallet as I milled the impulse face.  

Having seen many Synchronomes over the years, I think that there IS a great deal of in-built tolerance to the design which is a testament to the original idea.  Note, however, that very similar results can be obtained with a straight, flat impulse pallet - as Mr Parsons & Mr Ball found.

Cheers,
Ian R



-----Original Message-----
From: Chris <chris.b@...>
To: synchronome1@groups.io
Sent: Sat, 14 Nov 2020 7:24
Subject: Re: [synchronomeelectricclock] The Impulse Roller & Shape of Pallet

Hello Michael and Welcome to the forum.

Let us not forget that the impulse curve is also rotated about radii struck from the center of the rod. Which makes me wonder how the factory shaped the impulse curve on the raw pallet casting. Perhaps they used a high speed, copy lathe using a pantograph mechanism?
My own [pre-gravity arm buffer] 1919/20 Synchronome had apparently been run for decades with the pallet badly skewed to the plane of the arc. Requiring careful work to reshape the impulse curve. I used the highest speed in my old S&B lathe and various grades of abrasive paper. Wrapped and glued tightly around a slightly undersized rod to reshape the impulse curve in 3-dimensions. It would have required much higher speed to allow a form tool to shape the impulse curve due to extreme chatter and the weakness of the original casting.

Only after re-assembly did I realise that I had altered the overall length of the pallet slightly. Requiring the pendulum be moved bodily to the left. With further adjustments to suit. It has been running steadily for at least 35 years in my care so my "botching" seems to have worked. Or perhaps Synchronomes can shrug off the worst offences that can be thrown at them. Like being run out of line for long enough to wear the original pallet curve askew. New pallet castings became available later but it would been even more difficult to copy the original exactly than to "repair" it.

Regards
Chris.B


Ian Richardson
 

Hi Bepi,

"Mr Parsons & Mr Ball" is a reference to the origins of Gent's "Pulsynetic" system, invented by Parsons & Ball.  Early Gent clocks carried the legend "Parson's Patent".

Cheers,
Ian



-----Original Message-----
From: Bepi <pepicima@...>
To: synchronome1@groups.io
Sent: Sat, 14 Nov 2020 13:42
Subject: Re: [synchronomeelectricclock] The Impulse Roller & Shape of Pallet

Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
--
Bepi


Bepi
 

Hal, sorry for adding to the confusion, let me make it worse now. In my pseudo-synchronome I control the clock-cycle phase of the impulse by changing a time delay starting from the previous period vertical crossing but it's clearer to speak about a phase angle, as in the angle of the tangent rule, p.-

On Nov 14, 2020, at 13:46, H Hal via groups.io <haal@...> wrote:

hi Bepi,


can you explain the difference between phase and time in relation to impulse or are you using the terms interchangeably?


regards


hal

On 14/11/2020 12:42, Bepi wrote:
Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
--
Bepi


--
Bepi


Neville Michie
 

The phase of the impulse can be adjusted by moving the pendulum sideways with the adjustment
above the suspension. I seem to remember instructions for adjusting this so the roller was
halfway down the slope when the pendulum was still and the latch tripped.
cheers,
Neville Michie

On 14 Nov 2020, at 23:42, Bepi <pepicima@gmail.com> wrote:

Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners.
--
Bepi


H Hal
 

Don't worry Bepi,

I live in a constant state of confusion.

You are impulsing the pendulum at some point during its swing. There are some points (during the swing) that the impulse will be very useful and efficient OR other points where the impulse will be detrimental (the child pushed by dad on the swing gets higher and higher OR gets lower and lower to a stop because it is tea time).

The phase angle represents a point somewhere on that sine curve (which represents the position of the pendulum in time) where impulse may be given.

?


On 14/11/2020 13:11, Bepi wrote:
Hal, sorry for adding to the confusion, let me make it worse now. In my pseudo-synchronome I control the clock-cycle phase of the impulse by changing a time delay starting from the previous period vertical crossing but it's clearer to speak about a phase angle, as in the angle of the tangent rule, p.-

On Nov 14, 2020, at 13:46, H Hal via groups.io <haal@...> wrote:

hi Bepi,


can you explain the difference between phase and time in relation to impulse or are you using the terms interchangeably?


regards


hal

On 14/11/2020 12:42, Bepi wrote:
Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
--
Bepi


--
Bepi


John Haine
 

Hello Michael, nice to see you over here!  And thanks for the pictures.  Whether Hope-Jones actually tried to achieve a raised-cosine impulse profile is unclear, and Shortt's derivation seems to have been done during a talk by H-J (at the IEE I think) at which the latter was presumably describing the existing clock.  From what I can see in various publications, production pallets actually just machined the face to a circular profile, and somewhere there is a time-lapse video that shows the wheel impulsing where it actually lifts off the pallet around half way down the ramp, so what the impulse profile is for the standard synchronome I don't know.  My own version uses a stepper to lift and lower the arm and the timings are selected to make sure that the roller drops on the dead roll before the ramp and is lifted after.  The profile follows Shortt's prescription. 

You are right that you can cut the pallet using a cutter of the same diameter as the roller following the required path, however a complication is that the roller diameter doesn't correspond to standard end mills, so I generated the profile by computing the coordinates that cut the required curve corrected to use a 4 mm cutter.  (I got the chassis and mechanism of a Synchronome minus its pendulum on eBay.)  There are some pics here.  My pallet is flat, I just take care to line it up in the plane of swing.

Actually I am not convinced that the profile works as it is claimed.  When I started running my clock I set it up with the roller half way down the ramp with the pendulum stationary (moving the pendulum suspension to set).  I also had a "startup" routine that gave an impulse every 2 seconds to get the pendulum swinging then switched to maintaining mode.  However when I looked carefully at the timing, I discovered that the impulse gave a distinct phase advance which amounted to about 1 ms each time, which indicates that the impulse is early.  Eventually to counteract this I shifted the pendulum suspension to remove the phase shift, at which point the roller was no longer on the ramp at the centre position.  I haven't been able to explain this but I wonder if Shortt's theory doesn't take account of the dynamics of impulsing but is just a static view.

John.


Bepi
 

Hal, to minimize the escapement error, in my synchronome it can be made undetectable, you need to carefully center the impulse around the phase corresponding to a vertical pendulum rod. Adding energy when it's all kinetic doesn't change the period in the small angle approximation.
This is why I was looking for a way to mechanically control the phase with precision. Thanks Neville about your answer, I am still looking for a way to do it with better accuracy than by hand, I guess I need to devise a simple process to do it.
I am not claiming it's a needed adjustment for best accuracy, the opposite, I think I reported long ago in this forum that minimizing the escapement error doesn't significantly affect the period in the short turn accuracy, even a large escapement error seems to be very repeatable in a synchronome.



On Nov 14, 2020, at 14:47, H Hal via groups.io <haal@...> wrote:

Don't worry Bepi,

I live in a constant state of confusion.

You are impulsing the pendulum at some point during its swing. There are some points (during the swing) that the impulse will be very useful and efficient OR other points where the impulse will be detrimental (the child pushed by dad on the swing gets higher and higher OR gets lower and lower to a stop because it is tea time).

The phase angle represents a point somewhere on that sine curve (which represents the position of the pendulum in time) where impulse may be given.

?


On 14/11/2020 13:11, Bepi wrote:
Hal, sorry for adding to the confusion, let me make it worse now. In my pseudo-synchronome I control the clock-cycle phase of the impulse by changing a time delay starting from the previous period vertical crossing but it's clearer to speak about a phase angle, as in the angle of the tangent rule, p.-

On Nov 14, 2020, at 13:46, H Hal via groups.io <haal@...> wrote:

hi Bepi,


can you explain the difference between phase and time in relation to impulse or are you using the terms interchangeably?


regards


hal

On 14/11/2020 12:42, Bepi wrote:
Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
--
Bepi


--
Bepi


--
Bepi


Ian Richardson
 

Hi All,

To add even further to the confusion, in Shortt's "free pendulum" clock, the impulse arrangement is the other way round!  The pallet (actually the flat surface of a Brocot jewel) rolls over and around the wheel.  It starts "dead" as a tangent to the wheel at the top, and falls down, rotating the wheel, until it drops off.  That has always seemed to me to be a better arrangement which could easily have been incorporated into an otherwise standard Synchronome.  Has anyone ever done that?

Ian R



-----Original Message-----
From: Bepi <pepicima@...>
To: synchronome1@groups.io
Sent: Sat, 14 Nov 2020 16:02
Subject: Re: [synchronomeelectricclock] The Impulse Roller & Shape of Pallet

Hal, to minimize the escapement error, in my synchronome it can be made undetectable, you need to carefully center the impulse around the phase corresponding to a vertical pendulum rod. Adding energy when it's all kinetic doesn't change the period in the small angle approximation.
This is why I was looking for a way to mechanically control the phase with precision. Thanks Neville about your answer, I am still looking for a way to do it with better accuracy than by hand, I guess I need to devise a simple process to do it.
I am not claiming it's a needed adjustment for best accuracy, the opposite, I think I reported long ago in this forum that minimizing the escapement error doesn't significantly affect the period in the short turn accuracy, even a large escapement error seems to be very repeatable in a synchronome.



On Nov 14, 2020, at 14:47, H Hal via groups.io <haal@...> wrote:

Don't worry Bepi,
I live in a constant state of confusion.
You are impulsing the pendulum at some point during its swing. There are some points (during the swing) that the impulse will be very useful and efficient OR other points where the impulse will be detrimental (the child pushed by dad on the swing gets higher and higher OR gets lower and lower to a stop because it is tea time).
The phase angle represents a point somewhere on that sine curve (which represents the position of the pendulum in time) where impulse may be given.
?

On 14/11/2020 13:11, Bepi wrote:
Hal, sorry for adding to the confusion, let me make it worse now. In my pseudo-synchronome I control the clock-cycle phase of the impulse by changing a time delay starting from the previous period vertical crossing but it's clearer to speak about a phase angle, as in the angle of the tangent rule, p.-

On Nov 14, 2020, at 13:46, H Hal via groups.io <haal@...> wrote:

hi Bepi,

can you explain the difference between phase and time in relation to impulse or are you using the terms interchangeably?

regards

hal
On 14/11/2020 12:42, Bepi wrote:
Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
--
Bepi

--
Bepi

--
Bepi


Ian Richardson
 

I should have added that the "pallet" (ie. Brocot jewel) is attached to the gravity arm and the wheel is attached (albeit loosely) to the pendulum rod.

Ian


-----Original Message-----
From: Ian Richardson via groups.io <irichar361@...>
To: synchronome1@groups.io <synchronome1@groups.io>
Sent: Sat, 14 Nov 2020 17:02
Subject: Re: [synchronomeelectricclock] The Impulse Roller & Shape of Pallet

Hi All,

To add even further to the confusion, in Shortt's "free pendulum" clock, the impulse arrangement is the other way round!  The pallet (actually the flat surface of a Brocot jewel) rolls over and around the wheel.  It starts "dead" as a tangent to the wheel at the top, and falls down, rotating the wheel, until it drops off.  That has always seemed to me to be a better arrangement which could easily have been incorporated into an otherwise standard Synchronome.  Has anyone ever done that?

Ian R



-----Original Message-----
From: Bepi <pepicima@...>
To: synchronome1@groups.io
Sent: Sat, 14 Nov 2020 16:02
Subject: Re: [synchronomeelectricclock] The Impulse Roller & Shape of Pallet

Hal, to minimize the escapement error, in my synchronome it can be made undetectable, you need to carefully center the impulse around the phase corresponding to a vertical pendulum rod. Adding energy when it's all kinetic doesn't change the period in the small angle approximation.
This is why I was looking for a way to mechanically control the phase with precision. Thanks Neville about your answer, I am still looking for a way to do it with better accuracy than by hand, I guess I need to devise a simple process to do it.
I am not claiming it's a needed adjustment for best accuracy, the opposite, I think I reported long ago in this forum that minimizing the escapement error doesn't significantly affect the period in the short turn accuracy, even a large escapement error seems to be very repeatable in a synchronome.



On Nov 14, 2020, at 14:47, H Hal via groups.io <haal@...> wrote:

Don't worry Bepi,
I live in a constant state of confusion.
You are impulsing the pendulum at some point during its swing. There are some points (during the swing) that the impulse will be very useful and efficient OR other points where the impulse will be detrimental (the child pushed by dad on the swing gets higher and higher OR gets lower and lower to a stop because it is tea time).
The phase angle represents a point somewhere on that sine curve (which represents the position of the pendulum in time) where impulse may be given.
?

On 14/11/2020 13:11, Bepi wrote:
Hal, sorry for adding to the confusion, let me make it worse now. In my pseudo-synchronome I control the clock-cycle phase of the impulse by changing a time delay starting from the previous period vertical crossing but it's clearer to speak about a phase angle, as in the angle of the tangent rule, p.-

On Nov 14, 2020, at 13:46, H Hal via groups.io <haal@...> wrote:

hi Bepi,

can you explain the difference between phase and time in relation to impulse or are you using the terms interchangeably?

regards

hal
On 14/11/2020 12:42, Bepi wrote:
Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
--
Bepi

--
Bepi

--
Bepi


Odell, Edward
 

Hi Ian

 

Yes, that was done as a development before the Shortt in two types of clock, using the inertia bar principle.  In these the idea, as in the Shortt, is that the speed of pendulum movement lengthens or shortens the impulse duration to maintain constant amplitude.

 

This was first done in 1 second and ¾ second clocks but only 2 I know of survive.  Most unusually for Synchronome this mechanisms required a crutch and these clocks never made it into Synchronome brochures despite at least one having been sold to a customer.  See Bob miles book page 155.  I am very interested in this type if anyone knows of any more examples, I have one

 

The principle was then developed in the Shortt Observatory regulator

https://clockdoc.org/default.aspx?aid=11206

and nice paper in AHS journal March 2017

 

Eddy Odell

 

From: synchronome1@groups.io <synchronome1@groups.io> On Behalf Of Ian Richardson via groups.io
Sent: 14 November 2020 16:02
To: synchronome1@groups.io
Subject: Re: [synchronomeelectricclock] The Impulse Roller & Shape of Pallet

 

Hi All,

 

To add even further to the confusion, in Shortt's "free pendulum" clock, the impulse arrangement is the other way round!  The pallet (actually the flat surface of a Brocot jewel) rolls over and around the wheel.  It starts "dead" as a tangent to the wheel at the top, and falls down, rotating the wheel, until it drops off.  That has always seemed to me to be a better arrangement which could easily have been incorporated into an otherwise standard Synchronome.  Has anyone ever done that?

 

Ian R



-----Original Message-----
From: Bepi <pepicima@...>
To: synchronome1@groups.io
Sent: Sat, 14 Nov 2020 16:02
Subject: Re: [synchronomeelectricclock] The Impulse Roller & Shape of Pallet

Hal, to minimize the escapement error, in my synchronome it can be made undetectable, you need to carefully center the impulse around the phase corresponding to a vertical pendulum rod. Adding energy when it's all kinetic doesn't change the period in the small angle approximation.

This is why I was looking for a way to mechanically control the phase with precision. Thanks Neville about your answer, I am still looking for a way to do it with better accuracy than by hand, I guess I need to devise a simple process to do it.

I am not claiming it's a needed adjustment for best accuracy, the opposite, I think I reported long ago in this forum that minimizing the escapement error doesn't significantly affect the period in the short turn accuracy, even a large escapement error seems to be very repeatable in a synchronome.

 

 

 

On Nov 14, 2020, at 14:47, H Hal via groups.io <haal@...> wrote:

 

Don't worry Bepi,

I live in a constant state of confusion.

You are impulsing the pendulum at some point during its swing. There are some points (during the swing) that the impulse will be very useful and efficient OR other points where the impulse will be detrimental (the child pushed by dad on the swing gets higher and higher OR gets lower and lower to a stop because it is tea time).

The phase angle represents a point somewhere on that sine curve (which represents the position of the pendulum in time) where impulse may be given.

?

 

On 14/11/2020 13:11, Bepi wrote:

Hal, sorry for adding to the confusion, let me make it worse now. In my pseudo-synchronome I control the clock-cycle phase of the impulse by changing a time delay starting from the previous period vertical crossing but it's clearer to speak about a phase angle, as in the angle of the tangent rule, p.-



On Nov 14, 2020, at 13:46, H Hal via groups.io <haal@...> wrote:

 

hi Bepi,

 

can you explain the difference between phase and time in relation to impulse or are you using the terms interchangeably?

 

regards

 

hal

On 14/11/2020 12:42, Bepi wrote:

Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
--
Bepi

--
Bepi


--
Bepi


Michael Gilligan
 
Edited

Thanks for the welcome and the additional information, John
... I signed-up here on your recommendation at Model Engineer 

I should perhaps just mention, for everyone’s benefit, that my plots relate specifically to the illustration and description in ‘Electric Clocks’ and are probably not directly relevant to later production.

MichaelG.
.

When describing the Synchronome system at the Institution of Electrical Engineers in February, 1910, I drew a force curve and showed how the end of the pallet J should be shaped to achieve this. In the discussion which followed that paper, Mr. W. H. Shortt, M.Inst. C.E., contributed a mathematical solution of its best form, which I give with the assistance of fig. 101.


Bepi
 

Neville, I don't understand: when one moves the pendulum sideways the drawing arm unlocks the gravity arm with a phase which changes with continuity, it's equivalent to lengthening or shortening the drawing arm. At the same time the gravity arm would hit the pallet profile at different locations though, changing effectively the shape and strength of the pulse. Moving the pendulum to the left, from one position on, the pulse doesn't even change its timing at all.
I am confused, is there a way to cleanly change just the phase? Rotating gravity by tilting the whole support plate around the roller axis?

p.-

The phase of the impulse can be adjusted by moving the pendulum sideways with the adjustment
above the suspension.
I would be convinced


On Nov 14, 2020, at 14:20, Neville Michie via groups.io <neville_michie=yahoo.com@groups.io> wrote:

The phase of the impulse can be adjusted by moving the pendulum sideways with the adjustment
above the suspension. I seem to remember instructions for adjusting this so the roller was
halfway down the slope when the pendulum was still and the latch tripped.
cheers,
Neville Michie
On 14 Nov 2020, at 23:42, Bepi <pepicima@gmail.com> wrote:

Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners.
--
Bepi






--
Bepi


Neville Michie
 

If you were prepared to make a serious task out of the experiment,
you could add a horizontal bar with an adjustable weight, to displace the 
pendulum zero position. The escapement would stay the same, but the 
pendulum zero would shift. The major inconvenience is retuning 
the pendulum with the additional hardware. Se diagram.

cheers, Nevville

On 15 Nov 2020, at 06:53, Bepi <pepicima@...> wrote:

Neville, I don't understand: when one moves the pendulum sideways the drawing arm unlocks the gravity arm with a phase which changes with continuity, it's equivalent to lengthening or shortening the drawing arm. At the same time the gravity arm would hit the pallet profile at different locations though, changing effectively the shape and strength of the pulse. Moving the pendulum to the left, from one position on, the pulse doesn't even change its timing at all.
I am confused, is there a way to cleanly change just the phase? Rotating gravity by tilting the whole support plate around the roller axis?

p.-

The phase of the impulse can be adjusted by moving the pendulum sideways with the adjustment 
above the suspension.
I would be convinced 


On Nov 14, 2020, at 14:20, Neville Michie via groups.io <neville_michie@...> wrote:

The phase of the impulse can be adjusted by moving the pendulum sideways with the adjustment 
above the suspension. I seem to remember instructions for adjusting this so the roller was 
halfway down the slope when the pendulum was still and the latch tripped.
cheers, 
Neville Michie
On 14 Nov 2020, at 23:42, Bepi <pepicima@...> wrote:

Ian who are Mr Parsons and Mr Ball and where can I read about their results? I agree that the shape is probably of little importance, more relevant is the location of the impulse in phase/time.
Does anybody know how to regulate the impulse phase in a synchronome in an easy, controllable and accurate way?
The shape might be of some relevance in terms of repeatability which boils down to absence of sharp corners. 
-- 
Bepi










-- 
Bepi