Anomoly or Problem?

[Pip]

Hi all, newbie tinkerer here again with another question...

I bought a Seiko Bellmatic 4006-6011 from May 1973 and received it yesterday. All seems to be in working order, so today I thought I'd wear it and record the accuracy against the clock on the Watchville app on my phone. (I've found that the time on this app matches that on time.is pretty much bang on, so seems good enough for my purposes). This morning I did the figure of 8 winding for 40 seconds then set it and noted it as 31 seconds fast and I recorded the difference against the app five times today. Tonight I made a plot of the change and was surprised to find that the accuracy improved through the day. Below is a basic plot with the actual in red and the projected in blue. 

I've no idea why it would change like this, I honestly thought it would be the other way round (although thinking about it I'm not sure why other than general entropy). I'd love to hear if this is:

1. Normal;
2. Indicates a particular problem(s);
3. It's likely I just cocked up my readings;
4. Something I've not thought of.

This is a birth month/year watch for a family member so is a bit special and I want it to be as good as I can get it so would be grateful for any thoughts guys and girls.

Thanks,

Pip

[Pip]

Update to the graph... interesting!

 

[clockboy]

Watches run/perform at different rates depending on its position. That is why when regulating a watch it is adjusted in various positions (Dial up/dial down etc.) to get a good mean average. I suspect this is the reason you are getting differences..
Also as the mainspring winds down it will gradually loose power & hence slow down. However if it is an auto then in theory this should not happen. 

[Pip]

2 hours ago, clockboy said:

Watches run/perform at different rates depending on its position. That is why when regulating a watch it is adjusted in various positions (Dial up/dial down etc.) to get a good mean average. I suspect this is the reason you are getting differences..
Also as the mainspring winds down it will gradually loose power & hence slow down. However if it is an auto then in theory this should not happen. 

Great point Clockboy, I'd not thought about physical orientation and how that will affect the mechanical works. These readings were taken during the day whilst wearing it. Out and about all morning then office in afternoon and an hours drive followed by working at home on my laptop in the evening. I'll have to do several 24 hour tests in different positions to get a better feeling. Is there a standard way of doing this or just however one fancies?

The main thing I have taken from it is the thing needs a damn good service I guess?

[Brian3]

Different positions effect the balance Pip, the balance wheel will not be in perfect poise, thus giving the positional errors in most cases, (other factors do come into it), happens in all watches, we all strive to get the best possible balance to bring our error number (delta) to an acceptable level.

[Pip]

Just now, Brian3 said:

the balance wheel will not be in perfect poise

What?! It's only 43 and a half! My balance was in perfect poise at that age, making me into the fine athlete that I am today. Oh hang on, no - my balance went at about the same time as I started getting served in pubs!

Good point @Brian3, assuming the movement is original then there will be some wear in there as well. When I took the back off to have a quick look there were no service signatures so I can only guess it's not been regularly serviced if indeed at all. This might be one to send off to my man for a good overhaul - it's a bit too precious and I'm still practicing on cheaper pieces albeit I'm not generally destroying them now.

[JohnR725]

Out of curiosity when was the watch last serviced?  Then what sort of timekeeping are you hoping for?

So to understand what clockboy saying about regulating and timekeeping of your watch I have a link to a PDF that explains a lot of the terminology.

Then for proper timekeeping you need to time over 24 hours especially if the watch has a calendar mechanism that can affect timekeeping. Then other things that affect timekeeping like where was the watch when you are timing it? In other words were you wearing it in which case the automatic should've been able to keep the thing wound up. With a constant amplitude it should keep reasonably even time. If the watch was wound up and slowly unwinding the rate will change. If the watch hasn't been serviced in a while the rate of change will be more dramatic depending upon the condition of the lubricants and the cleanliness of the watch.

http://www.witschi.com/assets/files/sheets/Calculation of the values X-D-DVH-Di-Im-N.pdf

[Pip]

19 minutes ago, JohnR725 said:

Out of curiosity when was the watch last serviced?  Then what sort of timekeeping are you hoping for?

So to understand what clockboy saying about regulating and timekeeping of your watch I have a link to a PDF that explains a lot of the terminology.

Then for proper timekeeping you need to time over 24 hours especially if the watch has a calendar mechanism that can affect timekeeping. Then other things that affect timekeeping like where was the watch when you are timing it? In other words were you wearing it in which case the automatic should've been able to keep the thing wound up. With a constant amplitude it should keep reasonably even time. If the watch was wound up and slowly unwinding the rate will change. If the watch hasn't been serviced in a while the rate of change will be more dramatic depending upon the condition of the lubricants and the cleanliness of the watch.

http://www.witschi.com/assets/files/sheets/Calculation of the values X-D-DVH-Di-Im-N.pdf

Thanks for the good info John. I don't know if the watch has ever been serviced, there appear to be no marking on the inside of the case which I understand is where people often date/initial it. As the piece is 40+ years old I guess it must have been serviced at some point though or it may have packed up by now? (I don't know how long a Seiko would continue for without being lubricated?)

I don't have a timegrapher machine and it's not on my radar at the moment. I do have the cheap and basic Hairspring app on my phone which was recommended to me but I think I will download the Watch Tuner Timegrapher app which I read some decent things about, and at £10 I don't suppose I can go too wrong. 

[ftwizard]

I too have recently started in this watch fiddling game, and there are two items that I can't imagine living without, and they are, Timegrapher and Demagnetiser.

Without these you will end up chasing your tail all day long. Timegrapher first, is essential for general fiddling.

[JohnR725]

5 hours ago, Pip said:

Thanks for the good info John. I don't know if the watch has ever been serviced, there appear to be no marking on the inside of the case which I understand is where people often date/initial it.

Just because the inside of the watch isn't marked doesn't mean it hasn't been serviced. It depends upon the policy of whatever serviced it. It seems be the most common things I've seen are scratching variety of secret codes. Felt pen as it's not supposed to be permanent but recently frowned on because supposedly it admits fumes bad for the watch. Or some repair people do not mark the inside of the watch at all. So markings on the back are only a general indication perhaps.

One of the nice things with early organic lubricants is they would gum up with time. This means the watch would come to a stop before bad things happen. Modern lubricants chemically can do just about anything over time or possibly might last forever perhaps. Moebius who makes lubricants used to run interesting advertisements that there oil in this case synt-a-lube/9010. Does not gum up, does not spread and does not evaporate unaffected by atmospheric conditions so we can assume that basically last forever Although they don't actually say that. Strangely enough the new bottles you purchase have dates saying when it expires?

So for instance let's say your watch has never been serviced and Seiko used the same lubricants recommended in the service guide. So looking at the service guide they're using a mix of Moebius Oil in this case synt-a-lube/9010 their oil S-4 and a Moebius grease. Then as the common trend was using light lubricants where as today we would use heavier lubricants. So as far as lasting forever goes Moebius has a lubrication handbook with specifications and interesting information that applies to this discussion. So synt-a-lube/9010 is supposed to never evaporate but it can spread itself to the point of appearing to evaporate. I've noticed this in watches I've serviced over many years it just disappears. So the recommendation to prevent disappearing is surface treating which usually isn't mentioned in tech sheets which is why typically it's not done plus there is a cost factor.

So I was looking up the Watch Tuner Timegrapher app Looks like an interesting program. When you get that we can go back to testing your watch. Amplitude is a real clue of how well the watche is working and looks like timing app has this plus the beat of the watch. So wind the watch up then let it run 15 minutes to an hour to settle down. Usually recommendation to let the watch settle down when it's wound up especially on Non-automatic watches we can put that extra little bit of power which screws up timing results specifically too much power. So you want to time for 24 hours to prevent external influences from changing things you'll keep the watch in a stationary position except when you're timing it. Then for timing pick one position only to make things simple unless you're obsessed. Do not time in either dial up or dial down as they represent your least friction just pick a crown position Crown down 12 o'clock up something you remember. Then we can look at the before and after typically the watch companies time wound up and 24 hours later. But because you've got a nifty app there's nothing wrong with multiple timings during the 24 hours.

Then in the above paragraph Were taking a simple approach to timing the watch companies do time in multiple positions at the start and the end to figure out what's going on but I thought we can start with something simple.

 

 

[Pip]

So I was looking up the Watch Tuner Timegrapher app Looks like an interesting program. When you get that we can go back to testing your watch.

I downloaded it tonight, but spent the evening watching **BLEEP**nal smash Ludogorets and so haven't used it yet. Tomorrow I'll fire it up and see what happens.
Thanks for your superb post, excellent info!


Sent from my iPad using Tapatalk

[ftwizard]

**BLEEP**nal......

[Pip]

Haha! I can't believe I am unable to say the name of a footy team! Thankfully I'm not from S**BLEEP**horpe...! And if I lived in that famous town on the Isle of Man I don't know what I'd do!