Dynamic poising of a Sellita SW200-1 (ETA 2824)

[RCroman]

Hi,

I have an Oris watch based on the Sellita SW200-1 movement (a copy of the ETA 2824) that I am considering performing dynamic poising on. I'm fairly new to this, having done it on three low-end watches so far with reasonable success, following the procedures outlined here.

Here are the rates and amplitudes at low amplitude (~155°) and at full wind:

Position	Rate (low wind)	Amp	Rate (full wind)	Amp
DU	-26	166	-5	303
DD	-20	171	+4	301
PL	-20	152	-7	287
	-22	148	-10	281
PU	-20	155	-1	287
	-44	148	+3	284
PR	-39	160	+9	285
	-45	155	+8	289
PD	-39	151	+5	279
	-32	149	-2	280

 

Of course the first thing I want to do is to align the DU and DD rates. But the amplitudes are nearly identical, at least a full wind. What could be causing a nine-second difference in DU vs. DD with virtually no amplitude difference?

Thanks,

Russ

 

[jdm]

First, I'd say, disregard the values when unwound. Good timekeeping can be had on Swiss movements only above 220 deg - you can find various discussions on the subject here.

Then I can only suppose that the minimal difference between DU/DD is due to some end shake issue, where the balance and/or the escape wheel aligns in a very slightly different way. I'm sure this must be discussed in some advanced book or school under the chapter "positional adjustment", but wouldn't know where.

By the way, there is a section here where most people like to introduce themselves before asking questions.

[RCroman]

1 hour ago, jdm said:

First, I'd say, disregard the values when unwound. Good timekeeping can be had on Swiss movements only above 220 deg - you can find various discussions on the subject here.

Then I can only suppose that the minimal difference between DU/DD is due to some end shake issue, where the balance and/or the escape wheel aligns in a very slightly different way. I'm sure this must be discussed in some advanced book or school under the chapter "positional adjustment", but wouldn't know where.

By the way, there is a section here where most people like to introduce themselves before asking questions.

Thanks, jdm.

The low-amplitude measurements are purely for making it easier to determine where the heavy side of the balance wheel is, per the procedure discussed here. At higher amplitudes, the poise errors are less prominent and typically inverted (position that is slowest at low amplitude becomes the fastest at high amplitude).

I'll find the intro space and make an entrance.

[jdm]

34 minutes ago, RCroman said:

The low-amplitude measurements are purely for making it easier to determine where the heavy side of the balance wheel is, per the procedure discussed here. At higher amplitudes, the poise errors are less prominent and typically inverted (position that is slowest at low amplitude becomes the fastest at high amplitude).

You certainly know the subject better than me... good luck making it perfect   

[rodabod]

Those results at low amplitude already look excellent to me. I think it would be a tricky one to improve, as surely the 270 deg positional variation must be excellent.

[szbalogh]

There is a book Donald DeCarle: Practical watch adjusting. Stepy-by-step checks to make sure everything is ok which is adding to the rates.

These are not that bad results. But from a swiss movement we can get some more. 

Good amolitudes suggesting You can skip the mainspring and gear train part  

As JDM said, Du/DD deviance suggesting to much endshake. Jeweling tool needed here adjusting the whole setting depth.

Hairspring should be tinkered to proper vertical rates. Or You just adjust it to Your average hand (left or right, top or bottom on the wrist) positions. 

And what about the beat rate and the sound pattern? All ok? No other issue?

Edited October 15, 2017 by szbalogh

[vinn3]

good explanation. and referance.  vin

[RCroman]

On 10/15/2017 at 3:18 AM, szbalogh said:

There is a book Donald DeCarle: Practical watch adjusting. Stepy-by-step checks to make sure everything is ok which is adding to the rates.

These are not that bad results. But from a swiss movement we can get some more. 

Good amolitudes suggesting You can skip the mainspring and gear train part  

As JDM said, Du/DD deviance suggesting to much endshake. Jeweling tool needed here adjusting the whole setting depth.

Hairspring should be tinkered to proper vertical rates. Or You just adjust it to Your average hand (left or right, top or bottom on the wrist) positions. 

And what about the beat rate and the sound pattern? All ok? No other issue?

Thanks for the helpful reply! 

Beat error is good in all positions... 0.3ms or less.

I’ve got DeCarle’s book, so I’ll go through that step by step.

[JohnR725]

On 10/11/2017 at 7:26 AM, RCroman said:

But the amplitudes are nearly identical, at least a full wind. What could be causing a nine-second difference in DU vs. DD with virtually no amplitude difference?

One of the things that affects the results are the settings of the timing machine. For instance the averaging time if that set too short you going to see variations just because of that. Then how much time did you  allow the watch to stabilize between different positions that will change things.

Then to understand the timekeeping of this watch service manual is attached. Also attached some relevant images from the manual to see how you're doing compared to what the manual says you should be doing.

Then variations in timing? One of the things to try is run your watch through all the various positions write all the numbers down then repeat the entire procedure again see how close your numbers are. Watches themselves do a beautiful job of averaging over long periods of time. Timing machines only average over a short period of time.

 

 

SW200_1_33_20150226.pdf

[RCroman]

1 hour ago, JohnR725 said:

One of the things that affects the results are the settings of the timing machine. For instance the averaging time if that set too short you going to see variations just because of that. Then how much time did you  allow the watch to stabilize between different positions that will change things.

Then to understand the timekeeping of this watch service manual is attached. Also attached some relevant images from the manual to see how you're doing compared to what the manual says you should be doing.

Then variations in timing? One of the things to try is run your watch through all the various positions write all the numbers down then repeat the entire procedure again see how close your numbers are. Watches themselves do a beautiful job of averaging over long periods of time. Timing machines only average over a short period of time.

Thanks... especially for the service manual sections! Of course my goal is to beat the spec, or at least get well into the "premium" category.

I set the timing machine to 60-second averaging and waited about 20 seconds after changing positions to start each measurement. Repeated a couple of times... very consistent readings.

My main question at the moment is what could be causing a significant DU/DD rate difference with no change in amplitude. So far the suggested cause that seems the most likely is excessive balance end shake. It also occurred to me that possibly the hairspring is out of flat (or sitting low) such that it comes into contact with the ends of the regulator pins in the DD position, shortening the effective length and causing a fast rate. I guess excessive end shake could cause exactly that effect as well.

[jdm]

22 hours ago, RCroman said:

It also occurred to me that possibly the hairspring is out of flat (or sitting low) such that it comes into contact with the ends of the regulator pins in the DD position, shortening the effective length and causing a fast rate.

I don't think so. You would see it gaining minutes not seconds per day.

[JohnR725]

I might have missed it but reading your original posting did you tell us the condition of this watch? In other words is is a brand-new factory purchase or has it been in the field for a while? In order to dynamically poise a watch the watch itself basically has to be as perfect as it can be.

Then your regulator pin theory fits perfect with old pocket watches that typically have pins and they aren't always parallel. But usually with the nonparallel regulator pins the hairspring typically doesn't move that much it's usually not that big of a problem. Although it is always pointed out in the books to make sure your regulator pins are parallel. But as a reminder this is a modern watch a lot of the rules and thoughts the past don't apply anymore. The hairspring is as perfect as they could make it unless somebody's played with it. The regulator pins image attached appear to be the ETACHRON System. If you do a search in this discussion group you'll find it's come up in the past and their suggestions on how to adjust.