SKX007 +1 second off wrist -45 and very erratic on wrist

[BeepBoop]

Hi,

Hoping someone may have an idea what may be happening here. My SKX was running pretty well for 2 years generally at -20 per day but recently started running at -60 or more per day. I didn't notice at first as working from home it wasn't getting as charged or worn so presumed for the last month or so it was just a low charge. I don't remember a specific whack but it is my "beater" so may have got one.

Anyway once I tried giving it a charge and it still was dropping time. I opened her up and tried to regulate and have got it to about -1 per day when resting on it's back however when worn it becomes very erratic and can loose 45 seconds in a morning. Other times it's 15 seconds in a hour then stable. During this time I will often be typing small rather than vigorous movements. I do not believe the balance or spring was touched during the regulation and suspect I've just now seen the fault more clearly as I've been tracking time every few hours rather than few days.

Anyone have any ideas? I can remove the back again and take a macro shot of the spring\wheel if it would help.

Thanks for taking the time to read,

Andrew

[BeepBoop]

This is a macro of the movement, in case that is of help

[watchweasol]

I would take the back off and pay close attention to the action of the balance wheel/spring for its action to make sure its free and the coils are moving cleanly. Having said that its more than likely the case its is in need of a service/clean and oil and possibly a new mainspring as they do get tired, being constantly wound and re wound.

[BeepBoop]

Thanks for the reply watchweasol, I can't see anything obviously wrong but then I suspect I may not know what I am looking for. I've a attached a video of the movement in action.

[jdm]

2 hours ago, BeepBoop said:

Anyone have any ideas?

You need a timegrapher for a diagnosis, but the cost of that or any work is much higher than a new NH36 mov.t, about 22 GBP shipped. if your crown/stem is a single part you will also need a new thread crown to install it.

[Nucejoe]

A bit of history helps, how old is the watch?  and its service history? 

If only two years old then mainspring is unlikely to need replacement.

Cleaning the residual dirt all around the backplate with a brush before and as you are removing the backplate is a good pratice to make a habbit of, a spec of dirt finding its way inside the movement as you remove the backplate expectedly causes problems.

Apart from rate variation in different positions and just judging from your description , your watch could have easily got magnetized if you just put it on your smart phone for just a sec.

Regs 

Joe

 

 

 

 

 

 

[jdm]

Magnetization is not a consistent cause for a watch running slow or erratically, anyway since "anything is possible" a blind treatment with  GBP 5 demagnetizer box won' t hurt. The same device can always be useful to demagnetize screwdrivers and other tools.

[spectre6000]

I have a new NH38 movement in an old Seiko 5 case. It's the same as the ubiquitous NH35, but without any date function (the movement that replaced your 7S26C; it's substantially similar but with the addition of handwinding and hacking). Out of the box, it was within spec, but it's a generous spec, and I figured I could do better. I have a Chinese 1900 timegrapher, so I wanted to experiment a bit for myself. I wore it for a week like normal, and averaged the error out over the seven days (I forget where it started, and I didn't keep my notes). Then, I adjusted the error out accordingly only in the dial down position. Repeat until I got it where I was happy.

What I found was that there is a prolonged settling in between what I initially adjusted and where it settled, and where the same position would be a week later. It seems that some jostling helps. I would let it sit and settle on the timegrapher for however long, and it still seemed to wander a bit (in the slowerly direction) after I wore it for a few hours. At least that's my working theory... I could be missing a puzzle piece, and I'm sure someone will be along shortly to fill that in if that's the case.

I eventually got it to -2 and change seconds a day and called it good. This took several weeks, and is definitely NOT practical for a professional (who probably knows better anyway). To get there, I had to adjust it to +9 on the timegrapher. I haven't looked to see where that settled out to, or how the positional error breaks down. It has not yet been seven days in the -2-and-change adjustment, and as of right now it's -5 and change seconds after a few (4) hours shy of 3 days.

For comparison, I performed the same experiment with the same methodology at the same time on a ETA 2824-2, and it wanted precisely 0 dial down. It's settled out to -.25 to -.33 S/d. Currently at 4 seconds slow over 10 days, but it's morning and just sat stationary for 8 hours or so. 

I haven't paid super close attention to this yet, but anecdotally it seems like the ETA slows down overnight and catches up during the day, while the Seiko is the opposite. I'll start paying closer attention to this going forward.

Different movements are made to differing levels of precision. Lower precision is cheaper, but at the cost of accuracy. I feel like (as you seem to also feel) activity levels play a part. The Seiko is my "shop watch" for use working on cars, woodwork, doing things around the house, etc. The previous iteration survived all manner of hammering (shock), welding (magnetic fields), chemicals, and various other extremes of abuse without any noticeable degradation in performance. A car wreck (net 75 or so mph, head on, small overlap, I was in a classic car with no safety features, and my watch hand flew off the steering wheel and starred the windshield) eventually killed the watch... Or so I thought. The cause of death wasn't actual death, but a small black hair lodged in the escape wheel... A good service would have brought it back to life. Unfortunately, thinking it was completely wrecked inside, I took it apart early in my watchmaking explorations, and failed to get it back together without borking the hair spring. 

[BeepBoop]

Thanks for the replies, I'll source one of the blue box Chinese demagentisers and give that a go. As you say no harm in having the tool. The watch is 2 years old to me, never serviced or opened prior to large time loss starting and had been averaging a rough -20 p/d in that time. It was purchased gray market from ebay as new old stock but there looks to be an oil mark on the rotor so I suspect it has been opened at some point prior to my purchase. Just looking up the serial number it appears to be a 2015 manufacture.

I cannot be certain as I didn't pay close enough attention to the time loss prior to popping the case back. It may be that be prior to opening the back it was just dropping time and post regulating it has become erratic.

Whilst I wait for the demagnetiser to arrive, I'll give it a blow out with a rubber lens puffer in case there is something there, seal it up, let run out of charge then wear it to see if it settles over a number of days. If it's less erratic and on wrist after a few days I'll just regulate it to offset for this when off wrist. If nothing else if it's too erratic it'll be a fun a covid project to read up on NH36 replacements.

I've attached a phone timegraph readout but understand these are not great and my use of this may account for possible magnetisation based on the above comments.

[jdm]

Another possibility is that when worn it doesn't wind enough, runs almost out of power hence inaccurately.

That is where a cheap timegrapher comes very useful, it tells you right away the basic health of a watch. You may read that according to some is not a necessary tool for a beginner but the reality is thay for just 100 pounds there are now a lot of collectors and watch enthusiasts owning one. 

 

[spectre6000]

Magnetization can only make it run fast, not slow. Additionally, it won't be a little fast. What's happening is the coils of the hairspring stick together, and that effectively takes a huge amount of the spring out of play. It's like moving the regulating pins halfway around the spring.

I've not seen the phone timegrapher output, but assuming the triangles are recorded beats (clicking sounds really), unless you have a good explanation for those guys way out on the sides, you may check that your hairspring is in plane. It might be hitting the balance wheel on occasion. When it does that, it sucks energy out of the balance, and slows it down. Could be a lot of other things, but if I saw that on my timegrapher I'd be peeking in the side to see if I accidentally bumped the hairspring or something. Could be you just barely touched it while adjusting the regulator, and pushed it down a little between the pins. The pins on mine are a little tight, and I think that's why it takes a few hours of wearing for the watch to settle in (always slower, which would make sense with an effectively elongating hairspring).

[JohnR725]

16 hours ago, BeepBoop said:

 I opened her up and tried to regulate and have got it to about -1 per day when resting on it's back however when worn it becomes very erratic and can loose 45 seconds in a morning. Other times it's 15 seconds in a hour then stable.

I see you're using a phone app and it would be nice if you had a real timing machine. So are all your observations of erratic timekeeping based upon the application on the phone or some other method? In other words if you set the watch to a known times standard. Then look at the watch 24 hours later what sort of variation are you seeing?

 

Then a note regarding magnetism it can actually produce a sine wave effect. I've actually seen this before a really interesting sine wave.

[BeepBoop]

2 hours ago, JohnR725 said:

I see you're using a phone app and it would be nice if you had a real timing machine. So are all your observations of erratic timekeeping based upon the application on the phone or some other method? In other words if you set the watch to a known times standard. Then look at the watch 24 hours later what sort of variation are you seeing?

 

Hi JohnR, the erratic nature has not been seen on the phone app but I haven't used this much as it seems quite difficult to get a reading that isn't effected by background noise unless the case back is off. I have used it pre and post regulating to see roughly what the change did. So if it was reading -40 and I know I was -15 in wearing I'd adjust it until I was seeing a reading somewhere nearer -25 and then see how it behaved in real life.

In general timing is done recording the offset from a known time (time.is) during the day. In general this will show steady change off wrist, different with the watches position but positionally consistent. However when on wrist usually when typing, it can, but doesn't always drop say 45 seconds in a few hours. I've seen 3 or 4 seconds in 15 minutes then it will behave normally again.

Maybe this is normal and I've just never noticed when checking daily or weekly in the past and now I'm just over observing. 

 

[BeepBoop]

4 hours ago, spectre6000 said:

Magnetization can only make it run fast, not slow. Additionally, it won't be a little fast. 

I've not seen the phone timegrapher output, but assuming the triangles are recorded beats (clicking sounds really), unless you have a good explanation for those guys way out on the sides, you may check that your hairspring is in plane. It might be hitting the balance wheel on occasion. When it does that, it sucks energy out of the balance, and slows it down. Could be a lot of other things, but if I saw that on my timegrapher I'd be peeking in the side to see if I accidentally bumped the hairspring or something. Could be you just barely touched it while adjusting the regulator, and pushed it down a little between the pins. The pins on mine are a little tight, and I think that's why it takes a few hours of wearing for the watch to settle in (always slower, which would make sense with an effectively elongating hairspring).

Thanks for explaining magnetization, sounds like this is not likely in this case.

With the timegrapher, I find it very difficult to get a clean reading. Those errant readings appear to be outside noise. I get a lot more if I don't isolate the watch with a make shift sound booth (a mixing bowl inverted over the top).

I can't see anything wrong with the spring but don't really know what I'm looking at. I'll try and find a video of a correctly running one for comparison. But will also wear the watch for a few more days and see if the behavior settles or gives a reasonably consistent daily effect.

[jdm]

3 minutes ago, BeepBoop said:

With the timegrapher, I find it very difficult to get a clean reading.

Pedant note: a mobile phone can never be a timegrapher. A real albeit cheap machine doesn't have this problem, you place the watch on the stand fully closed and in any position, and it will read fine even if there is some other noise or music in the room. A sensitive but selective device.

 

[handwound]

2 minutes ago, BeepBoop said:

Thanks for explaining magnetization, sounds like this is not likely in this case.

With the timegrapher, I find it very difficult to get a clean reading. Those errant readings appear to be outside noise. I get a lot more if I don't isolate the watch with a make shift sound booth (a mixing bowl inverted over the top).

I can't see anything wrong with the spring but don't really know what I'm looking at. I'll try and find a video of a correctly running one for comparison. But will also wear the watch for a few more days and see if the behavior settles or gives a reasonably consistent daily effect.

When the balance is stopped check from the top if the coils of the hairspring are equally spaced (it's normal that the part of the spring where the regulator pin is attatched is further away than the other coils) and from the side if the hairspring is perfectly horizontal and not twisted in any way. If you can't observe anything abnormal here it is unlikely that the issue is caused by the hairspring. 

[jdm]

1 hour ago, handwound said:

When the balance is stopped check from the top if the coils of the hairspring are equally spaced (it's normal that the part of the spring where the regulator pin is attatched is further away than the other coils) and from the side if the hairspring is perfectly horizontal and not twisted in any way. If you can't observe anything abnormal here it is unlikely that the issue is caused by the hairspring. 

The problem is that for a beginner, sometime not having even an eyepiece, it's difficult to make these observations. And it's impossible to look by the side without removing the mov.t from case. Personally, after years, I still find difficult to ascertain what is going on under the cock in a 7S26.

Regarding the shape on the horizonal plane, that is best observed with the balance running. While it is normal that dog leg side "breaths" more, you will be looking for any spot where coils get dangerously close. Same for an even hairspring clearance within the regualator, is checked with the mov.t running. 

[handwound]

1 minute ago, jdm said:

The problem is that for a beginner, sometime not having even an eyepiece, it's difficult to make these observation. And it's impossible to look by the side do it without removing the mov.t from case. Personally after years I still find difficult to ascertain what is going on under the cock in a 7S26.

you are right, havn't thought about that  it really isn't that easy to spot a tiny kink in a hairspring and you need to know what to look for

[JohnR725]

20 hours ago, BeepBoop said:

specific whack but it is my "beater" so may have got one.

It's amazing what a whack can do to a watch especially if it was to move any of the regulation components.

16 hours ago, BeepBoop said:

The watch is 2 years old to me

The problem with how long you've had the watch is it doesn't tell us how long it was sitting wherever it came from. It's amazing how long some stores have watches in stock.

1 hour ago, jdm said:

The problem is that for a beginner, sometime not having even an eyepiece, it's difficult to make these observation. And it's impossible to look by the side do it without removing the mov.t from case. Personally after years I still find difficult to ascertain what is going on under the cock in a 7S26.

Really need a decent timing machine the Chinese 1000 or the 1900 outstanding well worth the money. Because otherwise were just guessing but I can make a guess.

If for unknown reasons any of the regulation components have moved because after all this is a etachron system is always conceivable something has rotated or moved. If the hairspring isn't absolute perfectly centered in the regulating pins or they opened up. we can get some interesting timing especially because you cannot wind this watch up. The problem with automatic watches is how much movement to the need to wind up?  For instance Seiko OEM watches give us a clue about this. They recommend if your winding by using the weight you need an auto wind. Turning at a speed of 30 revolutions per minute you'll be all wound up in 60 minutes. This means it's very conceivable that you're just not wound up enough and low amplitudes really screw up timekeeping. We really need a decent timing machine or we can do a heck of a lot a guessing.

[BeepBoop]

5 hours ago, JohnR725 said:

It's amazing what a whack can do to a watch especially if it was to move any of the regulation components.

The problem with how long you've had the watch is it doesn't tell us how long it was sitting wherever it came from. It's amazing how long some stores have watches in stock.

...

Thanks John

6 hours ago, handwound said:

you are right, havn't thought about that  it really isn't that easy to spot a tiny kink in a hairspring and you need to know what to look for

thanks handwound

6 hours ago, jdm said:

The problem is that for a beginner, sometime not having even an eyepiece, it's difficult to make these observations. And it's impossible to look by the side without removing the mov.t from case. Personally, after years, I still find difficult to ascertain what is going on under the cock in a 7S26.

Regarding the shape on the horizonal plane, that is best observed with the balance running. While it is normal that dog leg side "breaths" more, you will be looking for any spot where coils get dangerously close. Same for an even hairspring clearance within the regualator, is checked with the mov.t running. 

Thanks JDM

 

Thank you all your help\time is much appreciated. To summarise 

1. I know very little, and should think carefully before touching watches  and that a real timegrapher is a key tool to successful tinkering.

2. It's likely not magnetised, given the behaviour of losing time.

3. It could be a fault with the hairspring, or debris. I will inspect using what I have whilst not a loop I think I can rig up a macro lens, extension tubes, a dslr and the TV as a view screen to get a close look at the spring to see if my untrained eye can see anything wierd.

4. Closely timing it today I can see the time swings on wrist are not as extreme but are present compared to off wrist. The difference today is instead of just winding the watch after regulation and then wearing as normal I made a point to wind (30 seconds circular motion on a horizontal plane, watch face up)  evening, morning and lunch. This seems to have reduced the amplitude of the wild swings. This makes me think the watch is not holding a charge as well as previously and or my normal activity has become to sedentary or likely a mix of both. When the watch is lower on power it seems more susceptible to the on wrist positional changes. This may be stupid and is only anecdotal given the sample size.

Thank you all for your help I couldn't have expected so much assistance and patience. I will continue to investigate this via trial and error to see if this is a power reserve issue, and watch is in need of a service. Given the cost of an NH65, stem and crown. If this seems to be the case I'll look into finding someone more skilled than I to swap out the movement. This seems practical given the service vs movement cost.

Andrew

[watchweasol]

Hi  For your information  the power reserve of the 7S26 = 43hrs and the Sll NH35 -= 41 hrs  Information for the Sll can be found here  https://calibercorner.com/seiko-caliber-nh35a/   and on the 7S26 series on the Ranfft website. google Ranfft  seiko 7S26 

[BeepBoop]

Just wanted to post up update in case anyone else with a similar issue ever stumbles across this and to thank all the members.

It appears the erratic behavior was due to the watch being maintained at a low charge, not due to being broken but due to my lack of movement and naivety of the state of charge. Others will know if this is normal but it seems reasonable.

If manually wound to ensure a good charge (shaken) the watch behaves consistently. If wound and worn with descent levels of movement the watch behaves consistently. In both cases when the watch is given a fair chance at being charged it maintains consistent behavior on wrist and for over 24 hours off wrist. At some point after 24 hours off wrist it starts to drift a little from it's usual rates and become more susceptible to positional variation as the charge diminishes. For me the watch will still run after 36 hours off wrist but has lost approx 45s during the last 12 hours of that 36.

Thankfully my escapades have not broken the watch, I know a little more than I did before and it's now regulated at +8 rather than -30 ish, but time will tell on this, and it has a new greased gasket.

TLDR - Check it's got petrol before you disassemble the engine, unless you wanted to see what the engine looked like.

[handwound]

Good to hear that everything works as it should, enjoy your watch